Introduction China has completed the most consequential rewrite of its Trademark Law in over a decade. On June 26, 2026, the 23rd Meeting of the Standing Committee of the Fourteenth National People's Congress adopted a comprehensive revision of the Trademark Law of the People's Republic of China — the fifth amendment since the law was first enacted in 1982, and the first substantive overhaul since the narrow 2019 revision. The revised law, comprising 87 articles across nine chapters (up from 73 articles in eight chapters under the outgoing law), will enter into force on January 1, 2027. Trademarks registered before that date remain valid. For brand owners, in-house counsel, and IP practitioners with China exposure, this is not a routine update. The revision touches registration standards, opposition timelines, well-known mark protection, damages calculations, and — perhaps most significantly — the treatment of bad-faith and speculative filings that have long troubled foreig...
Background
Ericsson initiated patent infringement proceedings against Micromax Informatics Ltd. and Mercury Electronics Ltd., alleging infringement of patents relating to speech codec technology — the combined use of a speech coder and speech decoder to compress a caller's voice signal, transmit it as coded speech frames over a radio link, and decode it at the receiving end. A notable feature of this technology is discontinuous transmission: during periods when the speaker is inactive, no coded speech frames are sent; instead, the transmitter periodically sends speech parameters sufficient for the decoder to generate "comfort noise" in place of silence.
According to Ericsson's own patent specifications, speech coders and decoders are conventionally built into radio transmitters and receivers respectively, working together to enable voice communication over a radio link. A mobile phone, in this framing, is simply a conventional radio communication device — a radio transmitter with a speech coder, paired with a radio receiver with a speech decoder.
The patents asserted in the litigation are examined below.
The Asserted Patents
Patent No. 203716 (IN/PCT/2001/00551/MUM) — "A Method and Apparatus for Preserving Perceptually Relevant Non-Speech Information in an Audio Signal"
This patent claims a method involving two steps: first, determining whether an audio signal constitutes speech or noise (a well-established, conventional determination), and second, determining whether the signal contains non-speech information that is perceptually relevant to a listener — a determination that may selectively override the first.
On analysis, the claimed method does no more than perform and override this first-stage determination; it does not specify how either determination is actually carried out. Given that the first determination (speech versus noise) is already prior art, the second determination — assessing relevance to a listener — appears to follow as an obvious next step, since the question of relevance naturally arises the moment a signal is classified as speech or noise. Further, "relevance" of non-speech information is inherently subjective and will vary between individual listeners, raising questions about the claim's definiteness.
The corresponding apparatus claim covers a device performing this method, but any conventional mobile phone already contains all the components described, including the handling of non-tangible elements such as the audio signal itself. On this basis, the apparatus claim appears to lack novelty and to be obvious for the same reasons as the method claim.
The Letters Patent Deed (LPD) for this patent was issued on January 8, 2007.
Patent No. 213723 (IN/PCT/2001/00552/MUM) — "Method and Apparatus for Generating Comfort Noise in a Speech Decoder"
Although the title suggests a distinct invention directed to comfort-noise generation, the claims of this patent are, in substance, identical to those granted under Patent No. 203716.
The LPD for this patent was issued on February 1, 2008.
The Discrepancy in Filing Records
The international filing date is recorded as November 12, 1999 in the Patent Office's e-register, but as November 8, 1999 in the LPD issued by the Patent Office. The corresponding PCT application, PCT/SE/1999/002073 (international filing date November 12, 1999), entered the national phase in respect of both Indian applications — 00551/MUM and 00552/MUM.
Outside India, the corresponding PCT national phase applications were filed twice only in South Africa, a jurisdiction that does not conduct substantial examination of patent applications. India, by contrast, does apply substantial examination — yet granted essentially identical claims across both Indian applications. A comparison with the original PCT claims as published shows that the claims granted in India are identical to those in the PCT application, and notably, none of them make any reference to comfort noise generation, despite this being the very feature Ericsson has asserted in the litigation.
This raises serious questions about how two applications with substantially identical claims came to be granted as separate patents, and more broadly, about the rigor of the examination conducted by the Patent Office in this matter.
Significance of Patent No. 213723 in the Litigation
Patent No. 213723 appears central to Ericsson's case. In its initial communication alleging infringement, Ericsson mapped this patent's claims (at that stage identified only by application number, not patent number) onto the AMR speech codec, asserting that the Micromax X332 model supported AMR speech and could successfully establish an AMR speech session. Notably, the claims referenced in that communication differ from the claims as they now appear on the Patent Office website, and from the claims as originally filed under the PCT. Other patents were asserted subsequently, but without any comparable claim-by-claim mapping.
The Underlying Technology
The technology is said to reduce bandwidth and space requirements, allowing a service provider to serve more subscribers concurrently. It also serves to suppress background noise at the caller's end so that the decoded signal at the receiving end is clearer. In one disclosed embodiment, the voice signal is divided into frames, noise frames are identified, and transmission of those specific frames is selectively withheld — conserving bandwidth and enabling the network to support a greater number of simultaneous users — while corresponding comfort-noise frames are generated at the receiving end.
Although framed as method and apparatus claims, these are, in substance, system claims describing a particular arrangement of components — not all of which belong to Ericsson. What the claims disclose is essentially a need (clearer, more efficient voice transmission) and a broad outline of how that need might be met (receive, code, decode); the specific implementation could reasonably be arrived at by a person skilled — or even semi-skilled — in the art. It merits closer examination whether the patents actually claim novel functioning of the coder and decoder themselves, particularly in light of Section 3(k) of the Indian Patents Act, which excludes algorithms and certain computer-related subject matter from patentability.
It is also notable that this technology would, in principle, allow a telecom operator to identify which subscribers are using it and prioritize them — accommodating more simultaneous users while subscribers without the technology wait or are deprioritized until network load decreases.
Open Questions Worth Examining
Who is the true beneficiary, and who should bear liability? The principal beneficiaries of this technology appear to be telecom service providers, yet they have not been joined as parties to this litigation — despite the fact that the claimed system requires, at minimum, two subscriber handsets (each equipped with a coder and decoder) and at least one network tower. If the system as claimed cannot function without the participation of the service provider's infrastructure, it is worth asking whether an infringement action can properly proceed against the handset manufacturer alone. Where a system or method can only be performed collectively, there is a reasonable argument that liability, too, should be assessed collectively — raising a potential question of non-joinder of necessary parties, though this must be balanced against the practical difficulty of compelling a patentee to sue a specific telecom operator when subscribers themselves choose their service provider.
Did Ericsson bypass the actual chip manufacturer? If Micromax is a bona fide purchaser of the relevant chips or circuit boards from a third-party manufacturer — and if Ericsson was aware that its technology (the coding/decoding algorithms being the essential component) was being manufactured and sold by that company, which is understood to have also been named as a defendant — the question arises whether Ericsson pursued the handset company instead of the component manufacturer because the royalty recoverable from a chip supplier (given the comparatively low cost of a chip) would be far smaller than the royalty recoverable on the sale price of a complete mobile phone.
Taken together, Micromax may have strong grounds to contest the assertion of these patents: as a legitimate downstream purchaser, it would face only marginal exposure had Ericsson proceeded solely against the chip supplier; instead, Ericsson appears to have pursued the larger commercial target to secure a larger royalty base. Two questions in particular merit scrutiny — first, whether the suit is bad for non-joinder of the telecom service provider as a necessary party, and second, whether patent infringement can properly be asserted against a handset manufacturer merely because certain of its devices contain a chip that, only in combination with the service provider's network infrastructure at both ends, forms the complete system actually claimed. On this view, the claimed system cannot function as a whole without coding and decoding capability being present simultaneously at both ends of the communication link, supported by the service provider's own hardware.
Procedural History
A single bench of the Delhi High Court passed an ex parte interim order against Micromax, authorizing seizure of documents from its offices relating to the sale and import of mobile phones using the technology in question. Micromax's appeal to the Division Bench was directed back to the Single Bench, with liberty to approach the Division Bench again only if not heard within 30 days, in accordance with the Code of Civil Procedure. Faced with the prospect of its sales being disrupted in the interim, Micromax entered into an interim royalty arrangement with Ericsson, reportedly at rates between 1.25% and 2% of the sale price. The Single Bench subsequently referred the matter to mediation, appointing Justice A.P. Shah (Retd.) as mediator.
Concluding Observations
Should this litigation proceed to a full trial, Micromax would have substantial grounds to challenge Ericsson's assertion of these eight patents — from the apparent overlap and duplication between the two lead patents, to the discrepancy between the claims mapped in Ericsson's initial infringement communication and those actually on record, to the broader questions of non-joinder and system-claim liability discussed above. A fully litigated outcome on these issues could well become a landmark decision on how system claims — and liability for their infringement — are evaluated under Indian patent law.
This article is intended for general informational purposes and reflects analysis and commentary on publicly reported litigation. It does not constitute legal advice, and readers should consult a qualified patent attorney for guidance on specific matters.
Comments
Post a Comment