Koninklijke Philips N.V. v Asustek Computer Incorporation & Ors [2017] EWCA Civ 1526

Case No A3/2016/3683[2017] EWCA Civ 1526
IN THE COURT OF APPEAL (CIVIL DIVISION)
[2016] EWHC 2220 (Pat)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
CHANCERY DIVISION
PATENTS COURT
Mr Justice Arnold
Venue Royal Courts of Justice, Strand, London, WC2A 2LLDate 11/10/2017
Before
LADY JUSTICE ARDEN DBELORD JUSTICE KITCHINLORD JUSTICE FLOYD
Between
KONINKLIJKE PHILIPS N.V.Claimant/(1) ASUSTEK COMPUTER INCORPORATIONDefendants(2) ASUSTEK (UK) LIMITEDDefendant(3) ASUS TECHNOLOGY PTE. LTDDefendant(4) HTC CORPORATIONDefendants/(5) HTC EUROPE CO. LTDAppellants
James Abrahams QC (instructed by Hogan Lovells International LLP) for AppellantsMark Vanhegan QC and Adam Gamsa (instructed by Bristows LLP) for RespondentHearing Hearing dates: 27-28 June 2017
Judgment
[1]This appeal raises issues of construction of a patent cross-licensing agreement containing a covenant not to assert patents against third parties. The fourth and fifth defendants (together “HTC”) were such third parties and claimed the protection of the covenant. They appeal from the decision of Arnold J dated 9 September 2016 and his consequent order whereby he held that the benefit of the covenant contained in an agreement dated 26 February 1998 between Qualcomm Inc (“Qualcomm”) on the one hand and the claimant (“Philips”), US Philips Corp. and Philips Consumer Communications LP (“PCC”) on the other (“the 1998 Agreement”) did not extend to HTC’s acts of alleged patent infringement complained of by Philips in these proceedings. The judge’s decision was made after the trial of a preliminary issue in patent infringement proceedings brought by Philips against HTC and three other defendants (“ASUS”). The preliminary issue does not concern ASUS. In common with many judgments in this technical area, acronyms are used to identify various bodies and standards. Although I will endeavour to introduce the acronyms where they first occur, Annex 1 to this judgment provides a glossary which I hope will assist the reader who is not familiar with the terms.[2]The three patents in suit (“the patents”) are said by Philips to relate to technology referred to as High Speed Packet Access (“HSPA”) used in so called 3.5G mobile telephone systems. Philips has declared to the European Telecommunications Standards Institute (“ETSI”) that the three patents are essential to the Universal Mobile Telecommunications System (“UMTS”), a telecommunications standard developed under its auspices. In other words, Philips contends that it is not possible to comply with the UMTS standard without infringing the patents. Philips therefore bases its allegation of infringement on the fact that the HTC equipment of which it complains is said by HTC to comply with UMTS, and must therefore infringe its essential patents. Part of HTC’s defence is based on their contention that they have the benefit of the covenant by Philips not to assert against licensees of Qualcomm any patent of Philips which is or is claimed to be essential to UMTS. The licence is said to be contained in clause 4.3 of the 1998 Agreement, which I will set out below. HTC have other defences which await trial, subject to the outcome of this preliminary issue.[3]Philips accepts that HTC are licensees of Qualcomm, and are in principle entitled to the benefit of the covenant not to sue even though they are not parties to the 1998 Agreement. Philips disputes, however, that the covenant extends to HTC’s acts of alleged infringement.[4]It is common ground that the 1998 Agreement is governed by the law of the state of California, in the United States of America. There was evidence of California law before the judge (although the witnesses were not cross-examined). The judge made extensive findings as to the relevant California law to which it will be necessary to refer in due course. None of those findings is challenged on this appeal.[5]Before coming to the terms of the 1998 Agreement, it is necessary to explain the various standards, their nomenclature and their history and then to summarise the negotiations which led to the 1998 Agreement. Standards and nomenclature Multiple access and access control

Standards and nomenclature

[6]In order to make efficient use of bandwidth, mobile telecommunication systems need to have a technique of allocating resources between the multiple mobile phones located within the reach of a base station. This is referred to as a multiple access scheme. Mobile telecommunications systems are often characterised by reference to their approach to multiple access. The principle techniques of relevance to this case are: i) time division multiple access (“TDMA”); ii) frequency division multiple access (“FDMA”), and orthogonal frequency division multiple access (“OFDMA”); and iii) code division multiple access (“CDMA”).[7]Some multiple access schemes do not rely on only one of these techniques. Thus, for example, in February 1998, ETSI’s 2G digital cellular standard, GSM, used both TDMA and FDMA. Qualcomm were champions of CDMA. In the United States CDMA had been adopted in the interim standard known as IS-95. Qualcomm did not favour its use in combination with other multiple access techniques.[8]In addition to needing a multiple access scheme, mobile telecommunications systems employ mechanisms to control access to the network. Paging is a technique used on the downlink as part of access control regardless of the multiple access scheme. Random access is a technique used on the uplink, also as part of access control, again regardless of the multiple access technique. Both paging and random access allocate resources uniquely to individual users and necessarily employ an element of time division. However, the presence of these mechanisms would not affect the way in which the standard itself was characterised.

The UMTS Standard

[9]The UMTS 3G standard was being developed by ETSI at the time of the negotiations which led to the 1998 Agreement. In December 1997 ETSI published four proposals for the radio interface of the UMTS standard. Each proposal was developed by a "concept group" and the proposals themselves were each given a letter of the Greek alphabet. The Alpha concept was based on a wideband CDMA (W-CDMA) scheme, the Beta concept on a hybrid OFDMA/TDMA scheme, the Gamma concept on a wideband TDMA scheme, and the Delta concept on a hybrid TDMA/CDMA scheme.[10]At a meeting on 28 and 29 January 1998 an agreement was reached within ETSI that both the Alpha concept and the Delta concept would be taken forward. The Alpha concept (W-CDMA) was adopted for the proposed paired band (Frequency Division Duplex or FDD) mode of operation, and the Delta concept (TDMA/CDMA) would be adopted for what was proposed to be the unpaired band (Time Division Duplex or TDD) mode of operation. From this point forward a manufacturer of phones or base stations would have to decide whether to implement one or other of these modes, or both. It is important to recognise, however, that any given phone would only need to operate one of these modes. Philips and Qualcomm were both members of ETSI and their representatives were at the relevant ETSI meeting where this decision was made.[11]Also in early 1998 it was envisaged that the proposed ETSI UMTS standard would be developed further in various ways. For example it was envisaged that it might include “opportunity driven multiple access” or ODMA, an intelligent relaying protocol which sits on top of the WCDMA radio system.[12]In UMTS, a system of different "Releases" is used to ensure a controlled progression of the defined standards. A mobile system can be constructed based on the set of all specifications which comprise a given Release. A Release differs from the previous Release by having added functionality introduced as a result of ongoing standardisation work within the groups. The first full release of UMTS was called "Release 99" (because it was intended to be released in 1999, although in fact it was released in March 2000). Release 99 incorporated a development of the Alpha concept for FDD and a development of the Delta concept for TDD. Release 99 was followed by "Release 4", then "Release 5" and so on. Later Releases are backwards compatible with earlier Releases, i.e. a network adhering to a newer Release continues to support all handsets conforming with older Releases, and handsets adhering to a newer Release continue to operate with networks conforming with older Releases. A phone manufacturer is therefore not obliged to adopt functionality incorporated into a new Release, although his phones may, presumably, become less competitive as a result.[13]In March 2000, as part of Release 99, the Alpha concept was functionally enhanced by the introduction of a DSCH traffic channel. DSCH is a shared channel using both TDMA and CDMA. There was evidence that there were difficulties with the technical implementation of DSCH.[14]I have already mentioned HSPA (High Speed Packet Access) in paragraph 2 above, but I need to explain that HSPA is itself an umbrella term, covering the application of that technology to the downlink and the uplink. In 2002, Release 5 of UMTS added High Speed Downlink Packet Access or HSDPA to the FDD mode of UMTS, enabling mobile phones to download data at high speeds. Release 6, in 2004, added High Speed Uplink Packet Access or HSUPA to the FDD mode of UMTS. HSDPA and HSUPA are together referred to as HSPA. The Philips patents in suit are said to be essential to HSPA. The accused HTC mobile telephones implement Releases 5 and 6 of UMTS, and therefore have HSPA.

The negotiation of the 1998 Agreement

[15]The 1998 Agreement is a cross-licence between Philips/PCC and Qualcomm. The judge summarised the negotiations leading to the 1998 Agreement at paragraphs 58 to 63 of his judgment. In the mid to late 1990s, Philips had a mobile telecommunications business manufacturing GSM handsets. Philips' strategy was to expand this business globally, with the aim of becoming one of the top three players with a global market share of more than 20%. Accordingly, it sought to expand into the USA, but it needed to obtain a licence from Qualcomm under the latter's CDMA patent portfolio in order to do so.[16]Philips owned a portfolio of patents which it claimed were essential to GSM. By the beginning of 1998, Philips had entered into 13 licence agreements covering GSM.[17]Qualcomm owned a number of patents protecting the CDMA system that it had developed, which it called cdmaOne. In 1993 the TIA, the US standards-setting body, had adopted an interim standard for mobile telecommunications called IS-95 based on the physical layer of cdmaOne. IS-95 was the dominant 2G standard in the USA in the period preceding the 1998 Agreement.[18]Qualcomm's strategy at the time was to support worldwide adoption of its CDMA technology instead of TDMA or GSM. By September 1997 Qualcomm had an established licensing program with 30 subscriber and 10 infrastructure licensees. Qualcomm considered that it held the core patents for cdmaOne. While other companies claimed that they held patents that were relevant to CDMA, Qualcomm's stance was that a licence under those patents was unnecessary to practise IS-95.[19]Following a period of negotiation between Philips and Qualcomm beginning in November 1995, and in order to settle litigation in the US that had arisen between them in May 1997 concerning three Philips patents which Philips claimed to be essential to IS-95, Philips and Qualcomm entered into the 1998 Agreement. The 1998 Agreement began life as a Qualcomm form of agreement, but was negotiated between the parties.

The judge summarised Philips’ attitude to the negotiations and the agreement as follows:

“Philips insisted on an exclusion of TDMA technology from the scope of the 1998 Agreement. Philips viewed the 1998 Agreement as very one-sided in favour of Qualcomm, not least because no value was apparently attributed to Philips' patent portfolio. Philips considered that its TDMA patents should be excluded because the patent positions between Qualcomm and Philips were reversed in respect of TDMA technology as compared to CDMA technology.”
[21]At the instigation of Qualcomm, the wording "the proposed ETSI UMTS standard" was introduced into the definition of "CDMA Wireless Industry Standard" in the 1998 Agreement. This occurred very late in the course of the negotiations. It was not the focus of the 1998 Agreement.

The relevant terms of the 1998 Agreement

[22]Clause 4.3 provides (with added underlining and some irrelevant matter excluded):
"QUALCOMM Option to Obtain Covenant Not to Assert. Philips hereby grants QUALCOMM an option to designate any or all of its existing and future CDMA licensees as a 'CDMA Technically Necessary Patent Beneficiary.' QUALCOMM may exercise its option at any time and from time to time throughout the term of this Agreement, each time by sending written notice to Philips identifying each new CDMA Technically Necessary Patent Beneficiary. Philips, on behalf of itself and its Affiliates, hereby covenants that, as to each CDMA Technically Necessary Patent Beneficiary, Philips and its Affiliates will not assert any of their CDMA Technically Necessary Patents against any manufacture, use, sale, importation of equipment and/or components, or other acts of infringement, relating to a CDMA Wireless Industry Standard of a CDMA Technically Necessary Patent Beneficiary; [first, irrelevant proviso]; provided, however, that nothing in this Section 4.3 It is agreed that this is an obvious typographical error for 4.3 shall prohibit or otherwise limit Philips' right to assert any of its patents against any entity for infringement relating to any TDMA equipment or system (including, without limitation GSM, IS-54, PCS-1800, and PCS-1900)."
[23]CDMA Wireless Industry Standard is defined in clause 1 as follows:
"'CDMA Wireless Industry Standard' means standards for public code division multiple access communications including but not limited to IS-95A, IS-96A, IS-127, ANSI J-STD-008, the proposed ETSI UMTS standard, their subsequent releases, revisions and derivations, and any local and regional standards based substantially thereon, any wireless local loop or wireless PBX (private branch exchange) systems based substantially thereon, and the Globalstar Satellite System. For the purposes of this Agreement, including but not limited to determining whether a patent is a CDMA Technically Necessary Patent, CDMA Wireless Industry Standard includes all of the above-described standards and systems but does not include the GSM standard or any other standard which utilizes a TDMA over-the-air interface."
[24]Whilst the focus of this appeal is on clause 4.3 and the definition of CDMA Wireless Industry Standard, the background is the cross-licence between Philips/PCC and Qualcomm. The grants to Philips and PCC respectively were not in identical terms. As to PCC, the licence related, firstly, to PCC Licensed Products which meant “(1) Subscriber Units, (2) and complete wireless end user terminals … which include the ability to initiate and/or receive Wireless Communications in accordance with a CDMAWireless Industry Standard and (3) CDMA Modem Cards”. Secondly the licence to PCC related to CDMA ASICs. An ASIC is an application specific integrated circuit or chip. The licence to Philips was limited to “Components” which was defined to include electronic devices (including ASICs) designed for use in Wireless Applications conforming to a CDMA Wireless Industry Standard but with certain exclusions including CDMA ASICs and PCC Licensed Products. If Philips wanted a licence for CDMA ASICS, clause 4.1 gave it an option to obtain one, but the terms of clause 4.1 purported to prevent Philips from passing on to its customers the right to use any ASIC which infringed a Qualcomm patent.[25]Clause3.1.1 and3.1.2 thus provide:
“3.1.1 Subject to the terms and conditions of this Agreement, including but not limited to timely payment of the royalties set forth herein, QUALCOMM grants to PCC a personal, nontransferable, worldwide and nonexclusive royalty-bearing license under QUALCOMM’s Patents solely for Wireless Applications to (a) make (and have made), import, use and sell, lease or otherwise dispose of PCC Licensed Products, and (b) to make (and have made) CDMA ASICs and Components (provided that the material features of such CDMA ASICs have been designed by or for PCC and which designs are owned and used exclusively by PCC) and import, use and sell, lease and otherwise dispose of CDMA ASICs and Components but only if such CDMA ASICs and Components are included as part of and within complete PCC Licensed Products Sold by PCC (or as replacement parts for PCC Licensed Products previously sold by PCC). 3.1.2 QUALCOMM hereby grants to Philips a personal, nontransferable, worldwide, non-exclusive, fully-paid and royalty free license under QUALCOMM’s Patents to make (and have made), import, use and sell, lease or otherwise dispose of Components.”
[26]Clause 3.2 provided for PCC to make royalty payments. Having set out certain rates, the clause goes on to say that a particular rate shall apply for products “which can be used to initiate and/or receive Wireless telecommunications transmissions in accordance with any proposed ETSI UMTS standard not backwards compatible with IS 95, or any subsequent releases, revisions, derivations or local regional standards-based substantially thereon, unless and until Qualcomm begins selling commercial quantities of [products] which can be used to initiate and/or receive Wireless telecommunications transmissions in accordance with such proposed ETSI UMTS standard”. In other words, a special royalty rate applies when UMTS is not backwards compatible with IS-95 until Qualcomm begin selling products which can be used in accordance with UMTS.[27]Clause 5.1 provides for the licence in the other direction from Philips to Qualcomm:
“Philips hereby grants to QUALCOMM a personal, nontransferable, worldwide, nonexclusive, fully-paid and royalty-free license under Philips' Patents solely for Wireless Applications to make (and have made), import, use and sell, lease or otherwise dispose of QUALCOMM Licensed Products. No other, further or different license is hereby granted or implied.”
[28]It is common ground that the defined term “CDMA Wireless Industry Standard” limits the scope of both the licence from Qualcomm to PCC under clause 3.1 and the licence from Philips to Qualcomm under clause 5.1. Both licences are limited to devices and components implementing such a standard.[29]Having summarised the terms of the 1998 Agreement, I must now set out the issues and the judge’s findings of California Law.

The issues

[30]HTC contend that Philips’ claims of patent infringement in these proceedings are barred by clause 4.3 because: i) HTC is a “CDMA licensee” of Qualcomm pursuant to a Subscriber Unit License Agreement dated 20 December 2000 (“the SULA”); ii) Qualcomm has designated HTC as a “CDMA Technically Necessary Patent Beneficiary” in writing to Philips; iii) The patents are “CDMA Technically Necessary Patents” because they have been claimed by Philips to be essential to the UMTS standard; iv) The UMTS standard is a “CDMA Wireless Industry Standard”; v) The 1998 Agreement is governed by Californian law, according to which a third party beneficiary such as HTC can enforce the covenant and deploy it as a defence to a claim for patent infringement.[31]Before the judge, Philips relied on four points as to why clause 4.3 did not extend to HTC’s acts of infringement, of which only three are now relevant: i) The alleged infringements arise from HTC’s implementation of HSPA in its 3.5G mobile telephone handsets. Philips contend that these acts are excluded from the scope of clause 4.3 because implementing HSPA is not “an act of infringement relating to a CDMA Wireless Industry Standard” as defined in clause 1 of the 1998 Agreement (“Issue 1”). ii) Clause 4.3 contains a proviso preserving Philips’ right to assert any of its patents against any entity for an act of infringement relating to any TDMA equipment or system. Philips contends that HSPA utilises a TDMA system and HTC’s devices are TDMA equipment within the meaning of the proviso (“Issue 2”). iii) Philips contends that the covenant not to sue extends only to CDMA Technically Necessary Patent Beneficiaries who are dealing in Qualcomm Licensed Products, such as integrated circuits manufactured and supplied by Qualcomm pursuant to the 1998 Agreement. This point only affects some of HTC’s chipsets, namely those it purchases from one of Qualcomm’s competitors, Mediatek (“Issue 4”). The judge’s findings of Californian law[32]I would summarise the judge’s findings of California law as follows: i) The goal of contractual interpretation is to discern the mutual intention of the parties as at the date of the contract. ii) Intention is determined from objective manifestations of the parties’ intent both in the language of the contract and in extrinsic evidence. I would add that we were told that it was common ground that, as in this jurisdiction, evidence of the parties’ actual, subjective intentions is not relevant. iii) Where there is a dispute about the meaning of contractual language, there is a two-stage test for determining the correct construction. First the court asks whether the language is reasonably susceptible to the interpretation urged by each party. Secondly, if it is, the court asks what the parties intended the language to mean. iv) However broad the terms of the contract may be, the contract extends only to those things in respect of which it appears the parties intended to contract. Particular clauses of a contract are subordinate to this overall intent, so the court may imply into a term a limitation that is necessary to make the clause conform to the apparent object of the parties. v) Admissible extrinsic evidence includes: the circumstances under which the parties negotiated or entered into the contract; the object, nature and subject matter of the contract; how and why particular language emerged from the parties’ negotiations; and the subsequent conduct of the parties. vi) Extrinsic evidence may be used to aid interpretation, even if the contract language appears unambiguous on its face. Thus, extrinsic evidence may demonstrate that the contract is in fact reasonably susceptible to more than one interpretation. In theory, where the terms of the agreement are clear and explicit the language of the agreement will govern. In practice, however, the language of the agreement and extrinsic evidence are considered as a harmonious whole, both to ascertain whether there is an ambiguity and also so as to resolve any ambiguity. vii) The conduct of the parties following execution of the contract is admissible extrinsic evidence and is considered as one of the most reliable means of ascertaining the parties’ intent. This includes a party’s unilateral conduct, even if unknown to its counterparty, although in the latter situation less weight may be accorded to the evidence. Once a dispute arises as to the meaning of a contract, a party’s conduct is no longer probative of intent. viii) Interpretations that lead to absurd results are not favoured by the California courts. For example, it is an absurdity to suggest a reasonable business entity would contractually obligate itself to operate without regard to its business interests. ix) If the foregoing principles do not produce a clear answer, the Californian courts construe the contractual language against the party that supplied the language in question (a contra proferentem rule). The judge’s reasoning on Issues 1 and 2[33]In relation to Issue 1, the judge recorded the parties’ cases as follows: i) HTC’s case was that Philips claimed that the patents in suit were essential to (in one case) Release 5 and (in the other cases) to Release 6 of UMTS TS 25.214, which was a Technical Specification forming part of the UMTS standard. HSDPA was introduced in 2002 as part of Release 5 and HSUPA introduced in 2004 as part of Release 6, and these were “subsequent releases” of “the proposed ETSI UMTS standard” within the meaning the definition of CDMA Wireless Industry Standard in the 1998 Agreement. ii) Philips’ case was that the definition of CDMA Wireless Industry Standard in clause 1 was restricted to standards which specify “pure” CDMA, that is to say systems in which the primary means by which individual users’ traffic services are separated is code division alone, and excludes any system which utilises a TDMA over-the air interface.[34]The judge first considered the meaning of the opening words of the definition, namely that the defined term “means standards for public code division multiple access communications”. It was part of Philips’ case that this could only mean “pure” CDMA. The judge held that the wording permitted but did not compel the conclusion that hybrid standards were excluded. In any event the wording went on to say that a number of standards were included. The judge dealt with these in turn. Thus, IS-95 was the first revision of IS-95. Philips said that it was “pure” CDMA because the primary means by which individual users’ traffic services was separated was code division alone. HTC said it was a hybrid system because it employed time division on the paging and random access channels as a means of access control. The experts agreed that use of TDMA on these other channels “did not alter the fact that IS-95A was properly characterised as a CDMA system”. The further listed systems, IS-96A and IS-97, were for use in conjunction with IS-95A and therefore “did not alter the characterisation of IS-95A”. Similarly, ANSI J-STD-008 was “based on Qualcomm’s CDMA”, and thus was “a CDMA system.”[35]As to “the proposed ETSI UMTS standard” which clause 4.3 cites as another example of a CDMA Wireless Industry Standard, the judge found that the Alpha concept was “properly characterised as a CDMA system”, and both ETSI and Qualcomm had so characterised it. On the other hand Philips accepted that the Delta concept was a hybrid TDMA/CDMA system, deliberately designed with TDMA elements so as to enable it to work closely with GSM.[36]Taking these matters into account, the judge considered that the more natural interpretation of “the proposed ETSI UMTS standard” was that it embraced both the Alpha and the Delta concepts since it had been agreed that both would be incorporated into UMTS. However, he did not consider this to be so clear that the words were not also reasonably susceptible of Philips’ interpretation, which was that it only covered the pure CDMA Alpha concept.[37]Next the judge tackled the words “subsequent releases, revisions and derivations”. HTC argued that even if “the proposed ETSI UMTS standard” was limited to the Alpha concept, Releases 5 and 6 were properly described as “subsequent releases” of that proposed standard. The judge thought that if the “the proposed ETSI UMTS standard” was limited to the Alpha concept, it cannot have been the parties’ intention to include a new over-the-air interface just because the standards body had included it in a subsequent release. On the other hand, if “the proposed ETSI UMTS standard” was intended to include a hybrid TDMA/CDMA system then it made sense to include HSPA, which is another hybrid TDMA/CDMA system as forming part of a “subsequent release”.[38]Finally

the judge focused on the closing words of the definition:

“but does not include the GSM standard or any other standard which utilizes a TDMA over-the-air interface.”
Because GSM is a hybrid TDMA/FDMA system, Philips relied on these words to say that other hybrid systems which included TDMA were excluded, such as the Delta concept. HTC argued that the 1998 Agreement divided standards into two groups, namely CDMA and TDMA by reference to their fundamental access technologies, even if not pure CDMA or pure TDMA.[39]The judge considered that the problem with HTC’s approach was that it was not clear how a hybrid system such as the Delta concept was to be categorised, since it used both CDMA and TDMA. He therefore concluded that the closing words favoured Philips’ interpretation, but the words were not so clear when read in the context of the 1998 Agreement as a whole not to be susceptible of HTC’s interpretation.[40]The judge also considered a number of aspects of the extrinsic evidence which bore on construction. There were four points which Philips relied on: i) The parties’ purposes in entering into the 1998 Agreement. Philips’ purpose was essentially to obtain a licence in respect of cdmaOne/IS-95, which was a pure CDMA system. Further, it wanted to exclude TDMA from the agreement. On the other hand Qualcomm’s strategy was to support worldwide adoption of its CDMA technology, instead of GSM or TDMA. As a result Qualcomm required counterparties to grant it rights in relation to cdmaOne/IS-95, but had no need to obtain rights in relation to TDMA or hybrid systems. ii) The late stage in the negotiations that the wording “the proposed ETSI UMTS standard” was included. iii) The fact that no royalties were payable to Philips, and no reduction in the royalties payable to Qualcomm under the 1998 Agreement. Philips said it was commercially unreal for the licence to extend beyond pure CDMA in those circumstances. iv) Subsequent conduct of Qualcomm. This consisted of certain statements in so-called White Papers published by Qualcomm summarising third-party (including Philips) intellectual property rights benefitting Qualcomm’s customers in January 2004 and January 2008.[41]The judge pointed out, in relation to the first point, that HTC contended that Philips would have wanted and needed a licence which covered UMTS whether it was a pure or hybrid system, and that Qualcomm’s interests extended beyond pure CDMA systems, and also included implementing UMTS. As to the second point, he noted that Philips accepted that this implied that Qualcomm saw a benefit in obtaining rights under the Philips CDMA portfolio relevant to UMTS, but submitted that this could be explained by the fact that the Alpha concept was a CDMA system. Overall, however, he considered that each of Philips’ four points, but particularly the third and fourth points, provided support for Philips’ interpretation of clause 4.3.[42]HTC also relied on extrinsic evidence. This consisted of the SULA. Clause 5.9 of the SULA contains a warranty by Qualcomm that Philips had covenanted not to assert any of Philips’ CDMA Technically Necessary Patents (as defined in the SULA) against HTC’s manufacture, use, sale or importation of Qualifying Subscriber Units solely for Wireless Applications. It can be seen from the definition of Wireless Applications, which in turn depends on the definition of Common Air Interface, that clause 5.9 extends to hybrid CDMA systems such as TD-SCDMA. The judge considered that, as a later agreement between different parties, clause 5.9 of the SULA carried little weight.[43]Finally, the judge considered the contra proferentem rule relied on by Philips. As the wording “proposed ETSI UMTS standard” was introduced by Qualcomm, the definition of CDMA Wireless Industry Standard and hence clause 4.3 should be narrowly interpreted as confined to pure CDMA standards. Further, it would be in Philips’ interests for the granting clauses 3.2 and 5.1 to be construed in the same way. In the case of clause 3.2, the narrow construction would lead to a lower royalty burden for Philips, and in the case of clause 5 .1, the narrow construction would reduce the ambit of Qualcomm’s royalty free cross license. Against this, HTC argued that it would be in Philips’ interests for the definition to be widely construed, since that would affect the breath of the licence granted by Qualcomm under clause 3.1.1. The judge was not convinced that any advantage to Philips of having a wider licence under QUALCOMM’s patents with respect to hybrid systems would outweigh the disadvantages identified by Philips.

The judge expressed his conclusion at paragraph 107 as follows:

“In my judgment the definition of CDMA Wireless Industry Standard is reasonably susceptible of both parties' interpretations. HTC's strongest points are that it expressly includes "the proposed ETSI UMTS standard" and its "subsequent releases, revisions and derivations" and that, by the date of the 1998 Agreement, it was known that UMTS would include the Delta concept, which was a hybrid system. On the other hand, the closing words of the definition indicate that standards which include TDMA are not covered. In these circumstances it is legitimate to turn to the extrinsic evidence to resolve the ambiguity. As discussed above, I consider that the extrinsic evidence supports Philips' case as to what the parties intended. I do not consider that it is necessary to resort to the contra proferentem rule; but if it is, I consider that this also supports Philips' case. Accordingly, I conclude that the definition of CDMA Wireless Industry Standard is to be construed in manner contended for by Philips, that is to say, as being restricted to pure CDMA standards (in the sense explained above). Since HSPA is not a pure CDMA system, the Patents are not covered by clause 4.3.”
[45]The judge dealt with Issue 2 comparatively briefly. He assumed for these purposes that, contrary to the conclusion which he had reached on Issue 1, HSPA was within the definition of CDMA Wireless Industry Standard. Philips contended that, even on that assumption, the proviso to clause 4.3 excluded HSPA because it was a TDMA system, albeit a hybrid one.[46]The judge noted that the words of the proviso which excluded “any TDMA equipment or system” could exclude only pure TDMA, or could exclude both pure TDMA and hybrid TDMA. As with the definition of CDMA Wireless Industry Standard, the judge went on to consider the specific examples given. GSM was a hybrid TDMA/FDMA system, and the same is true of the other systems listed. If the words “TDMA equipment or system” included hybrid TDMA/FDMA, they could not be restricted to pure TDMA, and might include other hybrids such as TDMA/CDMA systems.[47]The judge again relied on the extrinsic evidence which he had considered in relation to Issue 1. He thought that Philips had a clear commercial reason for excluding both pure and hybrid TDMA systems from the scope of clause 4.3 and Qualcomm’s focus was upon pure CDMA.[48]It followed that HTC lost on both Issues 1 and 2, and their reliance on the 1998 Agreement failed for both reasons. The appeal on Issues 1 and 2 HTC’s arguments on Issue 1 and 2

The appeal on Issues 1 and 2

[49]Mr James Abrahams QC, who argued this appeal on behalf of HTC, submitted that the definition of CDMA Wireless Industry Standard was completely clear. The definition said in two places that the proposed ETSI UMTS standard was within the definition: firstly when it identified the proposed standard within the list of standards included in the opening words, and then again when it said that it “included all the above-described standards”. When the judge said in paragraph 107 that the closing words indicated that standards which included TDMA were not covered, he was ignoring the second, emphatic affirmation in those closing words that the proposed ETSI UMTS standard was covered. The judge ought not to have held that there was any ambiguity in the language. In any event, all the extrinsic evidence supported HTC’s and not Philips’ construction, or was neutral.[50]Mr Abrahams also had an alternative argument. Even if Philips was correct, and the definition only specifically included ETSI’s UMTS concept Alpha, this still did not help Philips. This was because HSPA was incorporated in Releases 5 and 6 of UMTS, each of which is a “subsequent release, revision or derivation” of the proposed ETSIUMTS standard, even as narrowly defined.[51]The judge had been wrong to hold that the closing words of the definition supported Philips. Given the clarity with which the definition included the proposed ETSI UMTS standard, it was clear that the closing words were not seeking to countermand what had gone before. Rather these words were included because Philips had an active GSM licensing programme and did not want to enable Qualcomm’s licensees to avoid taking a GSM licence in reliance on this provision.[52]Mr Abrahams submitted that the closing words of the definition reflected the fact that the 1998 Agreement took a binary view of the world, dividing it into CDMA standards like IS-95 and UMTS on the one hand and TDMA standards like GSM on the other. Although there could be difficulties allocating hybrid systems to one camp or the other, that is not a problem when it came to whether the definition of CDMA Wireless Industry Standard covered the proposed ETSI UMTS standard, because the definition twice announced that it did. The concluding words of the definition could not be read as excluding the proposed ETSI UMTS standard when the definition had gone to so much trouble to make it clear that it fell into the CDMA camp. What the definition as a whole was seeking to do was to distinguish between the groups of standards like IS-95 and UMTS where Qualcomm claimed a strong patent position and those like GSM in which the position was reversed and Philips had a strong patent position. Thus when the closing words said “GSM or any other standard which uses a TDMA over-the-air interface”, the word “other” meant “other than all those which had gone previously (including UMTS) and was also to be interpreted as meaning “of the same kind as GSM”.[53]As to the commercial background, Mr Abrahams submitted that the judge had lost sight of important evidence which showed that the background changed significantly between 1995 when negotiations started and 1998 when the agreement was concluded. He relied on some evidence given by Mr Wickins, who was Philips’ technical expert, that in 1998 it would have been clear to the industry that the third generation of mobile telephony in the form of UMTS was going to be “a big thing” in the near future. More generally, anybody in the industry at that time who wanted to become a major player in mobile telephony would have to be making serious preparations for the forthcoming UMTS standard. Philips had by this stage, as the judge found, a plan to become a global player, one of the top three manufacturers in the world, with a 20% share of the global mobile telecoms market. Accordingly, submitted Mr Abrahams, Philips required a licence to do UMTS, whatever direction UMTS ultimately took. Qualcomm was asserting a strong position in respect of both the Alpha and Delta concepts: Philips would have been keen to obtain rights in respect of both technologies.[54]Mr Abrahams submitted that Qualcomm was a major player in mobile telephony and it was therefore equally important to it that it obtained all necessary rights from Philips to implement the forthcoming UMTS standard, so that it could make chipsets which implemented the standard. It would also need such a licence, again whatever the direction UMTS ultimately took. Thus Qualcomm had become a member of ETSI in late 1997. Moreover it was Qualcomm who sought to include UMTS within the scope of the 1998 Agreement. He relied upon a passage in Philips’ competition law pleading in the present case:
“Qualcomm was … aware by the time the 1998 Agreement was entered into that its WCDMA technology had been selected by ETSI to be incorporated into UMTS and so was seeking to develop a licensing strategy that would enable it to obtain a dominant position in UMTS Technology Markets and UMTS Chipset Merchant Markets.”
[55]Thus, submitted Mr Abrahams, whilst Qualcomm may have had a technical strategy of promoting CDMA, it was also a chipset manufacturer with an obvious interest in being free to manufacture chips which were compliant with UMTS, whichever direction it went in.[56]Mr Abrahams submitted that this commercial background was reflected in the terms of the 1998 Agreement itself. He relied on clause 3.2, which provided for a differential rate of royalty for Philips to pay depending on whether UMTS was backwards compatible with Qualcomm’s IS-95 technology. This showed, firstly, that the parties contemplated that UMTS might not be “pure” CDMA. Secondly it showed that Philips would still want to implement UMTS even if it was not pure CDMA, and therefore needed a licence for that purpose. Thirdly it showed that Qualcomm at least contemplated making chipsets in accordance with a UMTS standard which was not IS-95 compatible. None of this made sense if the 1998 Agreement was only ever going to give rights to make pure CDMA phones and chipsets.[57]As to the fact that UMTS was added late, Mr Abrahams submitted that this was wholly explained by the fact that UMTS was in development in the course of the negotiation of the 1998 Agreement, and that by 1998 both parties’ commercial strategies would have required the necessary licences under the other side’s standard-essential patents in order to participate in it. In fact, the stage at which the references to UMTS were introduced favoured HTC, because they showed that the words were added at a point in time when both parties knew that the proposed ETSI UMTS standard would include CDMA/TDMA.[58]The judge’s reliance on the absence of royalty payments to Philips was also explicable on the basis that the agreement was a cross-licence. Qualcomm had a large portfolio of relevant patents and Philips had a small portfolio. There was nothing significant therefore in the payment of royalties only being one-way. The evidence was that Qualcomm attached no value to Philips’ patents in any event. Ultimately Mr Abrahams submitted that the one-way nature of the royalty provisions was neutral on the meaning of the definition.[59]Mr Abrahams submitted that the judge had also misunderstood the message conveyed by Qualcomm to its customers in the White Papers published in 2004 and 2008. The White Papers were not saying that customers only acquired rights in relation to pure CDMA. Instead, they stated that the customers would get rights to Philips’ patents which were essential to “W-CDMA”. W-CDMA was the FDD mode of UMTS. By 2004, W-CDMA included two traffic channels which were hybrid TDMA/CDMA - the DSCH channel and HSDPA. By 2008, W-CDMA included HSUPA as well. In addition the White Papers contained language which pointed out that Qualcomm had additional rights from third parties (including Philips) with respect to other air-interface standards eg TD-SCDMA, which was a time division/code division hybrid multiple access system.[60]Finally, Mr Abrahams repeated the submission he made to the judge about the impact of the SULA between Qualcomm and HTC. Clause 5.9 was a warranty from Qualcomm that Philips had covenanted not to assert its Philips CDMA Technically Necessary Patents against customers of Qualcomm. The defined class of patents expressly included a hybrid TDMA/CDMA system, namely TD-SCDMA. This demonstrated Qualcomm’s intention and understanding that the 1998 Agreement extended to the whole of UMTS, including parts which were not CDMA. If the judge’s construction were correct, Qualcomm would be in breach of the warranty in clause 5.9 of the SULA, because they were not in a position to assure their customers that Philips had covenanted not to assert these patents. The judge had been wrong to reject HTC’s reliance on the SULA on the grounds that it was a later agreement between different parties, whilst at the same time apparently placing significant reliance on the White Papers, which also involved only Qualcomm and its customers, and were likely to be less well considered statements than those made by way of warranty in a formal agreement. Indeed the White Papers contained a disclaimer which indicated that Qualcomm was not guaranteeing that its interpretation of the third party rights was correct.[61]Mr Abrahams submitted that the contra proferentem rule also assisted HTC. Once the commercial background was taken into account, and it is clear that both sides wanted the right to practise UMTS, it made no sense to construe the 1998 Agreement narrowly so as to avoid that outcome.[62]As to Issue 2, Mr Abrahams submitted that the words of the proviso were a belt-and-braces provision, but they simply gave effect to the last sentence of the definition of CDMA Wireless Industry Standard. Both were inserted to protect Philips’ GSM licensing programme. Further, it made no technical sense for clause 4.3 not to have the same technical scope as the licence from Philips to Qualcomm. The overall intention of the clause included the fact that Qualcomm was licensed to make products such as chipsets and could then sell them to its customers so that the customers could use them in their own products and not be sued by Philips. If HTC were right under Issue 1, Qualcomm was licensed to make UMTS chipsets because UMTS is, on this assumption, a CDMA Wireless Industry Standard. It followed that Qualcomm’s customers were free to incorporate UMTS chipsets in their phones. Finally Mr Abrahams relied on his analysis of the extrinsic evidence, considered under Issue 1, as supporting HTC’s case and not that of Philips.

Philips’ arguments on Issues 1 and 2

[63]Mr Vanhegan QC, who appeared for Philips, submitted that HTC’s arguments on Issues 1 and 2 illegitimately painted a picture of the commercial background which was contrary to that accepted by the judge. At the date of the 1998 Agreement, Qualcomm still considered that UMTS, if it was going to be developed, should be a Qualcomm-based pure CDMA system. Moreover all Philips wanted was a licence to enter the US market with a 2G pure CDMA system. HTC’s alternative scenario was based on a hypothesis as to the parties’ real commercial objectives which was not tested by being put to Philips’ witness, Mr Amar.[64]Thus, in paragraph 45 of his witness statement, Mr Amar had said that, in finalising the 1998 Agreement, Philips had insisted on an exclusion of TDMA technology. He continued:
“46. This exclusion was added to the definition of CDMA Wireless Industry Standards as well as separately in the covenant not to assert in clause 4.3. I recall that the feeling at the time was that the agreement was very one-sided in favour of Qualcomm in the sense that the primary consideration for the licence to Qualcomm’s CDMA system was the royalty payments from Philips. Qualcomm did not attribute any value to Philips’ CDMA patents, as was reflected in the royalty rate to be applied … I recall at the time that there was a strong feeling in Philips that TDMA should be excluded because the patent positions as between Qualcomm and Phillips were reversed for TDMA. The 1998 agreement was regarded as strictly limited to CDMA only.”
[65]Mr Amar then explained the origin of the words “the proposed ETSI UMTS standard” in the definition, and the various Greek letter concept groups which existed. He continued:
“50. While Philips were involved in this process, the personnel directly engaged in the licensing negotiations with Qualcomm were not. We had a general awareness of what was going on and understood that one of the proposals was based on CDMA. Not long after the 1998 agreement was entered into, I recall that Qualcomm pushed to establish a single worldwide CDMA standard for 3G. Qualcomm’s stance on the choice of technology for 3G is recorded in the ITU press release… Qualcomm’s public position on a single worldwide CDMA standard as recorded by the ITU is consistent with my recollection of Qualcomm’s approach to the 1998 agreement as strictly limited to CDMA only.”
[66]Based on this evidence Mr Vanhegan submitted that because Qualcomm held all the cards on CDMA, it was simply not interested in acquiring rights to UMTS if it became a hybrid CDMA/TDMA standard.[67]Mr Vanhegan submitted that these paragraphs of Mr Amar’s statement were a legitimate foundation in California law for showing the parties respective interests and purposes in entering into the 1998 Agreement. He showed us the evidence of Professor Leftsin, a professor at Hastings College of the Law in San Francisco, which he submitted supported that approach.[68]So far as Qualcomm’s purpose was concerned, Mr Vanhegan relied further on two items of evidence: i) A response filed by Qualcomm at a meeting in March 1998, after the date of the agreement, commenting on the ETSI proposal for UMTS. Qualcomm’s position was that the proposals adopted at the January meeting (and in particular concept Alpha) were not as good as Qualcomm’s pure CDMA proposal. Going forward it was Qualcomm’s view that there should be only one proposal and that should be based on Qualcomm’s own proposal, rather than concept Alpha. ii) The ITU Press Release, referred to in paragraph 50 of Mr Amar’s statement, again post the agreement, in which Qualcomm stated that it was only prepared to licence its CDMA technology for UMTS if three conditions were met:(i) a single converged CDMA standard set for 3G;(ii) the converged standard accommodated equally the two dominant network standards;(iii) certain disputes on technological points were resolved.[69]In support of his argument that the agreement was a US-focused agreement to enable Philips to enter the US market for 2G products, Mr Vanhegan drew attention to the fact that all the parties apart from Philips were US entities. PCC was a joint venture limited partnership between Philips and Lucent Technologies Inc, itself a US entity. Although the joint venture obtained the right to enter the US market and elsewhere, Philips itself did not get an individual right to make basic CDMA equipment. Mr Vanhegan pointed to clause 3.1.2 which does not give Philips itself (as opposed to Philips’ joint venture with Lucent) rights in CDMA products.[70]Mr Vanhegan reiterated the conclusions of the judge, namely that the definition of CDMA Wireless Industry Standard included only standards which specify “pure” CDMA. He submitted that the judge’s conclusion that the clause was reasonably susceptible of this construction was correct, and that the extrinsic evidence and the language of the definition, as a harmonious whole, supported the judge’s narrow construction.[71]As to the extrinsic evidence, Mr Vanhegan stressed that, in 1998, the Alpha and Delta concepts were separate proposals for the FDD and TDD modes of operation. Any given telephone complying with the UMTS standard would have to offer one or other of those modes. Although a dual mode phone was a possibility, there was no mandatory provision of the standard requiring both modes. Accordingly, in 1998, it was entirely reasonable for a manufacturer to envisage making phones in accordance with one or other concept. There was nothing technically unreasonable in the concept of a pure CDMA phone. Moreover the Alpha concept was perfectly capable of being a self-contained basis for standardisation.[72]The absence of a provision for payment of royalties to Philips was a highly material piece of extrinsic evidence. It did not make sense for Philips to be granting significant TDMA rights to Qualcomm when the royalty payments were all one way. Although it was a cross-licence, Philips were not granted any discount on Qualcomm’s royalty rates for CDMA: accordingly if there was a licence of TDMA technology it was one which Philips had given away.[73]Next Mr Vanhegan stressed that the language which introduced the ambiguity, namely the introduction of the example of the ETSI UMTS standard, was language which was introduced by Qualcomm. So the language did not support the notion that it was introduced in order to give Philips rights to UMTS. The wide construction would result in a wider licence from Philips to Qualcomm and a wider covenant not to assert against third parties under clause 4.3. It also has the potential for Philips to have to pay more royalties under clause 3.2. Thus the contra proferentem rule, if it was necessary to apply it, would have favoured Philips.[74]So far as the White Papers were concerned, Mr Vanhegan submitted that there was a clear distinction between pure CDMA (which the White Papers called W-CDMA) on the one hand and hybrid TDMA/CDMA on the other (in the form of TD-SCDMA). In the former case the customer is protected, but in the latter case the customer is told that he will need to discuss with Qualcomm whether they had rights. HTC’s argument that W-CDMA had by this stage developed to a point where it included TDMA aspects (such as DSCH and HSPA) was not argued before the judge.[75]HTC’s argument that W-CDMA was widely defined was incorrect for a number of reasons, including the fact that, based on a textbook before the judge, all W-CDMA as implemented was pure CDMA.[76]In relation to the SULA Mr Vanhegan submitted that it was in Qualcomm’s interest to give the widest possible meaning to the definition of the royalty bearing unit, to maximise its royalty return against HTC. However, when it came to the warranty contained in clause 5.9 of the SULA, it was made clear that Philips retained its rights to assert its patents against HTC for infringement relating “to any time division multiple access (TDMA) equipment or system (including, without limitation GSM, IS-54, the PCS-1800, and PCS-1900).” The wider definition of CDMA (which included hybrids) did not therefore ultimately assist HTC.[77]Turning to Issue 2, Mr Vanhegan drew attention to clause 4.1 of the 1998 Agreement, the clause which provides Philips with an option for a licence to make CDMA ASICs, but which expressly does not enable Philips to confer rights on its customers. That, he submitted, refuted HTC’s point that the covenant not to sue in clause 4.3 must have the same technical scope as the rights which are cross-licensed between Philips and Qualcomm. Here was a provision, said Mr Vanhegan which placed an obligation on Philips’ customers to obtain a licence from Qualcomm, notwithstanding the fact that Philips had the right to make the relevant chips.[78]Mr Vanhegan submitted that the proviso to clause 4.3 did not relate to standards in general but related to particular acts of implementation by third parties. Standards frequently offer alternative modes of operation. Thus one did not have to have both FDD and TDD in the same phone. Further, HSPA is an option to releases 5 and 6 of UMTS. The judge had made a finding to that effect in paragraph 116 of his judgment when dealing with the issue of whether the Philips patents in suit were essential to the UMTS standard. Clause 4.3 gave the beneficiary the right to implement any release of UMTS so long as the beneficiary’s particular implementation did not include TDMA technology. The proviso did not prevent the beneficiary from complying with the standards in general; it simply limited the extent to which that beneficiary could use all the possible technology which is set out in that standard. HSPA used TDMA as the primary access technology. Devices which chose to implement HSPA were therefore excluded from the covenant contained in clause 4.3.

Discussion on Issues 1 and 2

[79]Because California law takes a holistic attitude to interpretation of a contract, it is, in my judgment important to consider Issues 1 and 2 together. It was both sides’ primary case that both the carve-out from the definition of CDMA Wireless Industry Standard and the proviso to clause 4.3 had the same underlying purpose, although they were divided on what that purpose was. HTC’s case was that the purpose was to protect Philips’ GSM licensing programme, whilst Philips’ case was that it was to protect Philips’ stronger position in TDMA generally.[80]Although Mr Abrahams submitted that the language of the definition of CDMA Wireless Industry Standard was not reasonably susceptible to Philips’ interpretation, that submission cannot limit the extent of the debate in the present case, because California law does not halt the process of interpretation at that point. An ambiguity may appear from the extrinsic evidence. As the judge put it, the language of the agreement and the extrinsic evidence are considered as a harmonious whole both to ascertain whether there is an ambiguity and also to resolve it.[81]Certain aspects of the definition are tolerably clear. It is hardly necessary to point out that the opening words define a category of standards. The definition contains examples of standards which are included (e.g. IS-95) and standards which are excluded (e.g. GSM). To be included, the standard must be within the general language “for public code division multiple access communication”, that is to say the standard must be one for public communication, and it must employ CDMA. There is nothing in the general words to suggest that they are intended to exclude standards which employ additional multiple access techniques as well. The definition then goes on to list four standards and the proposed ETSI UMTS standard as all being included as non-limiting examples. It makes clear that these standards are included in the definition together with their subsequent releases, revisions and derivations, thereby making it clear that the definition is not concerned with the precise technical content of the listed standard at the date of the agreement, although how much deviation is permitted is not specified either. The listed standards would, however cease to be within the definition if they ceased to employ code division multiple access.[82]The concluding words of the definition begin by reinforcing the impression gained from the opening words, namely that all the previously listed standards and systems are included. However it continues by saying that GSM, or any other standard which uses a TDMA over the air interface is excluded. There is a potential difficulty, therefore, as to whether a hybrid CDMA/TDMA system is included by the opening words (because it is “for public code division multiple access communication”) or whether it is excluded by the closing words (because it “utilizes a TDMA over-the-air interface”).[83]One approach to the resolution of this difficulty is that taken by HTC. That is to say that the drafter has resolved the dilemma in the case of the proposed ETSI UMTS standard by stating expressly in the opening words that it is included and reinforcing that conclusion in the closing words. Once that is appreciated, there is no difficulty in understanding that other standards involving CDMA (i.e. other than the list in the opening words) which also utilise a TDMA over-the air-interface are excluded.[84]Philips’ alternative construction relies heavily on the reader extracting the common characteristic “pure CDMA” from the list of standards expressly included. Even if the list had not expressly included the proposed ETSI UMTS standard, this would be a difficult argument. Philips’ attempt to restrict CDMA to “pure CDMA” cannot be achieved by reliance on the closing words (excluding any system which utilises a TDMA over-the-air interface). The specific exclusion of TDMA is not sufficient for Philips’ purposes, because the definition can only be restricted to pure CDMA if the opening words inherently exclude any other multiple access system (e.g. FDMA) and not if it excludes just TDMA. Viewed the other way round, if the opening words are limited to pure CDMA, it would not be necessary to exclude systems which utilise TDMA by the express closing words.[85]Moreover, the fact that the exemplified standards conform to Philips’ “pure CDMA” definition does not show an intention to exclude from the definition systems which use CDMA together with other multiple access systems. They are all well known examples of systems which do use CDMA only. However, given their status as examples of systems which use CDMA, they do not compel a construction restricted to pure CDMA.[86]Once the proposed ETSI UMTS standard and its subsequent releases, revisions and derivations are expressly included in the list, Philips’ argument based on the language alone, at least to my mind, becomes yet more difficult. Firstly, one could scarcely ask for a clearer answer to the question “is the proposed ETSI UMTS standard (and its subsequent releases) within the definition CDMA Wireless Industry Standard?” The answer is obviously that they are, because it expressly says so. Secondly, Philips do not escape from this conclusion simply by pointing out that one concept within that proposed standard was pure CDMA, i.e. the Alpha concept. The language refers the reader to the proposed standard, not some part of it.[87]Save for one point, I would, for my part, have approached the extrinsic evidence on the basis that the language of the definition did not admit of Philips’ construction. However, California law makes clear that particular clauses of a contract are subordinate to the overall intent of the parties. It cannot therefore be right to construe the definition and clause 4.3 in hermetically sealed compartments. The proviso to clause 4.3 makes it clear that, at least for third parties, there is no protection for “any TDMA equipment or system”. That expression operates, at least in part, at the level of the individual equipment, rather than the standard, and makes it at least arguable that the carve-out from the definition was also for everything involving a TDMA over the air interface, even within one of the included standards. Even if the proviso does not cross-fertilise the meaning of the definition in this way, it may be the correct way to read the proviso. I think however that the carve-out from the definition is also reasonably susceptible to an interpretation which excludes from the definition any standard which utilises TDMA. That would exclude the proposed ETSI UMTS standard to the extent that it utilised a TDMA over the air interface, even if it did not support Philips’ more extreme “pure CDMA” construction.[88]The task for the court, applying California law, would then be to see whether the extrinsic evidence helped to resolve the issue between these two interpretations to which the language is reasonably susceptible.[89]I take first Philips’ deployment of the evidence of Mr Amar. Care is needed to isolate from it matters of objective fact which support Philips’ interpretation from Mr Amar’s recollection of his feelings and impressions. The latter are not admissible on the issue of construction under California law any more than they are under ours. Professor Lefstin made this clear in the passages of his report on which Mr Vanhegan relied.[90]However Prof Lefstin also made clear that the objective facts existing at the date of the agreement would, under California law, take into account the state of the relevant technology that forms the subject matter of the agreement, the state of the relevant regulatory environment, and the positions of the parties with respect to the relevant technology and regulatory environment. This would include the state of technology and standards as understood at the time of the agreement, and each party’s motivation in the context of its larger business posture and strategy. I would add that it must include the strength of the parties’ patent positions with regard to CDMA and TDMA.[91]I consider, next, Mr Wickins’ evidence (relied on by HTC) that it would have been clear to the industry in 1998 that the third generation of mobile telephony in the form of UMTS was going to be “a big thing” in the near future. More generally, anybody in the industry at that time who was or wanted to become a major player in mobile telephony would have to be making serious preparations for the forthcoming UMTS standard. This evidence is significant, but it must not be taken too far. It does not mean that the parties must have intended that they would enjoy licences to practise the UMTS standard however it developed. The evidence would, for example, preclude an interpretation which prevented any sensible implementation of UMTS at all. Because UMTS could be practised using concept Alpha, it cannot be assumed that it was intended that the licence would necessarily be any wider.[92]Equally, the fact that Philips was focused on obtaining a licence for the US is not, in my judgment, particularly persuasive. I accept that the occasion for the 1998 Agreement was Philips’ focus on the US market for 2G and IS-95. The scope of the 1998 Agreement extended further, however. There was no geographical limit on the rights which were granted, and the express inclusion of the ETSI UMTS standard proposal shows that the parties’ focus was not limited to the US but included other standards as well. Philips had a plan, as the judge found, to become a global player and one of the top three manufacturers in the world, with a 20% share of the global mobile telecoms market, which would not have been limited to IS-95.[93]It is therefore clear that Philips would have wanted a CDMA licence to practise the forthcoming UMTS standard. However, as I have said, it is not a necessary consequence of the natural desire for such a licence that Philips would have wanted to do so to the fullest extent possible, and in whatever direction the UMTS standard should go. A viable approach for Philips would have been to regard a licence which enabled them to practice the Alpha concept as adequate. That was a viable approach from the technical perspective. If Philips wanted to practise the Delta concept, or further TDMA features were added, further cross-licensing might be necessary.[94]Turning to Qualcomm, the first point to note is that it was Qualcomm who requested the inclusion of the proposed ETSI UMTS standard in the agreement. I accept that part of Mr Abrahams’ submission which relies on clause 3.2 as showing that Qualcomm had a real interest in participating in the UMTS standard, as any chipset manufacturer would wish to do. On the other hand, I am not persuaded that one can derive from that clause any express indication that Qualcomm contemplated participating in UMTS if it was not pure CDMA: purity of CDMA and backwards-compatibility are not the same thing. However, as with Philips, a desire to participate in UMTS does not necessarily imply a desire to do so to the fullest extent possible, and in whatever direction the UMTS standard should go. The Alpha concept was a viable technical proposal, and Qualcomm would have wished to use its strong patent position in CDMA to ensure that UMTS stayed so far as possible on the CDMA track.[95]There is some support for this in the post-agreement material relied on by Philips. Thus, Qualcomm’s statement at the March 1998 meeting was clearly part of the negotiating process which sought to steer the industry towards CDMA. The ITU press release is also a statement of Qualcomm’s negotiating position, saying that it was not prepared to license its IP rights at all (including for UMTS) unless there is a worldwide converged CDMA standard.[96]The judge recited both sides’ contentions on commercial purpose, and concluded that he considered that this head of evidence favoured Philips. To be fair to the judge, it was not one of the matters on which he relied in particular. To my mind, this head of extrinsic evidence does not clearly support Philips’ or HTC’s interpretation, but is not inconsistent with either.[97]The next item of extrinsic evidence relied on by the judge is the late stage in the negotiations at which the reference to the proposed ETSI UMTS standard was introduced. Again, the judge did not place particular weight on this head of extrinsic evidence. It supports his finding that UMTS was not the focus of the 1998 Agreement, and provides very modest support for the idea that the parties might have been content with less than comprehensive rights for UMTS.[98]The third item of extrinsic evidence identified by the judge, and the first on which he placed particular reliance, is the absence of an express provision for payment of royalties to Philips. To my mind this is of very considerable significance. The royalty provisions are structured so as to lead to the result that, if Philips was licensing its TDMA rights, it was doing so for nothing. Yet the evidence showed that Philips had a much stronger position in TDMA than did Qualcomm. That, to my mind, is a clear indicator that, if there is a practical and realistic construction of the agreement which does not result in Philips giving away its TDMA rights, then that construction should be adopted.[99]The fourth item of extrinsic evidence was the White Papers. The judge held that the White Papers assisted Philips. Despite Mr Abrahams’ ingenious argument that by the date of the White Papers the term W-CDMA included aspects of TDMA, I think that the judge was entitled to say that they supported a narrow construction of the licensed rights. The White Papers draw a clear distinction between the rights customers will obtain and those which they will not. In the former category were W-CDMA and cdma2000, but “other air interface standards” were not covered. I accept Mr Vanhegan’s submission that W-CDMA is likely to be taken in this context as excluding TDMA, although the extent to which this helps is somewhat counteracted by Qualcomm’s contrasting position in the SULA.[100]Standing back, the picture which emerges from the extrinsic evidence after the full argument which took place before us, is not as clear to me as it appeared to the judge. Nevertheless, the fact that I feel much less confident than the judge did about some of the items of extrinsic evidence does not mean that I would be justified in reaching a different conclusion on this appeal. None of the evidence mandates HTC’s construction and the absence of a royalty is a powerful indicator in favour of Philips.[101]I turn to Issue 2 and the proviso to clause 4.3. I have already said something about the language used in the proviso, which seems to me to exclude all TDMA equipment from the operation of the covenant. The difference in language cannot be attributed to the different context of clause 4.3, which is directed at “acts of infringement … relating to a CDMA Wireless Industry Standard”. If one were following the approach used in the definition, the carve-out from this language which the proviso needed to effect would be by reference to acts of infringement relating to other standards. However the proviso goes further and excludes acts of infringement relating to TDMA equipment.[102]To my mind, when this clause is considered with the extrinsic evidence the position is clear. Philips had not granted Qualcomm rights in respect of TDMA equipment and were not covenanting not to sue in respect of it. That is strongly supported by the absence of a royalty in favour of Philips. Equipment which implements HSPA is TDMA equipment. Accordingly clause 4.3 does not provide HTC with a defence.[103]Viewing both issues 1 and 2 together, I am not persuaded that the judge reached an incorrect conclusion. I would dismiss the appeal on issues 1 and 2.[104]I do not think that HTC’s alternative argument succeeds either. The alternative argument would be that the later Releases of UMTS are subsequent releases of the ETSI UMTS proposal. If “the proposed ETSI UMTS standard” excludes the optional concept Delta because it utilises a TDMA over the air interface, then it also excludes subsequent optional functionality which uses such an interface. The relevant equipment is, in any event, excluded by the proviso to clause 4.3. The judge’s reasoning on Issue 4[105]Issue 4 concerns whether clause 4.3 extends to chipsets purchased from third parties. The judge dealt with this at paragraph 121-123 of his judgment. He accepted Philips’ case, namely that it would be very surprising if the parties had intended third parties to benefit from the equivalent of a royalty-free licence from Philips in respect of products which did not contain any components supplied by Qualcomm, as that would be contrary to both parties’ commercial interests. He also concluded that this interpretation was supported by extrinsic evidence, namely the post-contractual conduct of the parties, evidenced by the terms of the notification letters under clause 4.3 sent out by Qualcomm to Philips between 2000 and August 2008, statements made by Qualcomm in the White Papers from 2004 to 2008, and the terms of the SULA. All these materials indicated that Qualcomm considered that clause 4.3 was limited to Qualcomm Licensed Products, and did not extend to products incorporating chipsets manufactured by third parties. The appeal on Issue 4 HTC’s arguments on Issue 4

The appeal on Issue 4

[106]Mr Abrahams submitted that there was no express language in clause 4.3 to indicate that it was limited to the use of chipsets manufactured by Qualcomm. If these sophisticated parties had wanted to limit the language of clause 4.3 in the way suggested by Philips, they could easily have done so.[107]The licence from Qualcomm to HTC, the SULA, expressly covers both phones containing Qualcomm chips and those not containing such chips. It was true that it only gave a warranty that Philips had covenanted not to sue in respect of the former, but that was not the same thing. It only showed that Qualcomm were not, for whatever reason prepared to grant a wider warranty, so as to cover products not incorporating their chips.[108]The judge had been wrong to say that it was contrary to both parties’ commercial interests for the licence to be as wide as HTC contended. There was obviously a benefit to Qualcomm in having the ability to provide a broader protection to its customers.[109]The notification letters had originally designated the customer only insofar as its products contained Qualcomm chipsets. However, this did not show that Qualcomm thought that clause 4.3 was limited in that way. Moreover Philips had responded by saying the letter was deficient, and that the 1998 Agreement did not give Qualcomm the right to limit its notification in that way. Mr Abrahams submitted that one can infer that Philips thought that clause 4.3 extended to all products.

Philips’ arguments on Issue 4

[110]Mr Vanhegan relied upon the decision of the Landgericht Mannheim in corresponding proceedings between the parties in Germany. That decision appears to support Philips’ case on this issue, albeit largely by reference to the judgment of Arnold J which we are considering on this appeal. The Mannheim decision does not appear to deal with issues 1 and 2 although they were argued before it. The decision is under appeal to the Oberlandesgricht. One of the grounds of appeal is that the court did not give any reasons for rejecting HTC’s case on issues 1 and 2, and Mr Vanhegan was not able to identify any reasoning on which he could rely. Whilst the decision of the Landgericht Mannheim is entitled to respect, it does not absolve us from considering Issue 4 for ourselves.[111]Mr Vanhegan submitted that if the covenant not to sue covered non-Qualcomm products, it would be a bad commercial outcome for both parties to the 1998 Agreement. The only possible beneficiary would be a person in the shoes of HTC who wished to buy from one of Qualcomm’s competitors. It was a commercial absurdity from the point of view of the parties to the 1998 Agreement. It would remove any commercial incentive for the customer to buy from Qualcomm. A construction which is contrary to the interests of both parties is an example of the kind of construction which California law treats as absurd.[112]The notification letters correctly reflected the scope of clause 4.3 as limited to the case where the equipment incorporates a Qualcomm ASIC. The letters dated after August 2008 were in any event inadmissible because a dispute had arisen between Qualcomm and Philips under the agreement. California law excludes reliance on subsequent conduct when a dispute has arisen.[113]The White Papers also supported Philips, as they contained numerous references to the fact that the beneficiaries only get the benefit if they acquire a Qualcomm ASIC. It was not possible to ignore these, as HTC sought to do, as the White Papers all asserted that they set out “the rights that QUALCOMM is able to provide to you…”. If HTC were correct, this statement about the rights which Qualcomm had available to grant would be, at the very least, disingenuous.

Discussion on Issue 4

[114]I can deal with this point very briefly as I agree with the judge for the reasons he gave. California law does not prohibit the court from implying a term or limitation to give effect to the parties’ intentions. The advantage to Qualcomm in being able to provide a wider covenant to its customers is not a real advantage as a matter of commercial reality. Qualcomm would not want to find ways of encouraging its customers to buy chips from one of its competitors. In reality, the result contended for by HTC is contrary to the interests of both parties to the agreement, as the judge found. The subsequent conduct relied on to demonstrate the parties’ intentions is, in this instance, persuasive.[115]I would therefore dismiss the appeal on Issue 4 as well.

Overall result

[116]I would dismiss the appeal.[117]I agree.[118]I also agree.