TQ Delta, LLC v ZYXEL Communications Ltd & Anor

In construing a patent claim, technical terms are interpreted purposively and in context. A term such as “frame” may have different meanings in different parts of the technology and is not necessarily a term of art with one fixed meaning. The scope of a patent monopoly must be justified by its technical contribution, but obviousness is assessed against the actual teaching of the prior art rather than by hindsight reconstruction.

Where a technical problem is routinely encountered in the relevant field, perceiving that problem is unlikely to involve an inventive step. A skilled person is deemed to read and assimilate pleaded prior art properly, including a lengthy incorporated standard, while focusing on the aspects relevant to the invention.

TQ Delta claimed that two DSL-related European patents were essential to relevant telecommunications standards. The first, EP 1 453 268, concerned controlling the transmission rate of overhead data by specifying which frames contained overhead bits. The second, EP 1 792 430, concerned normalising CRC anomaly counts where the CRC computation period varied.

The defendants challenged validity and infringement. The central issues were construction, obviousness over Alabama in relation to the 268 Patent, and obviousness over ADSL2/02 in relation to the 430 Patent. The court also considered whether the patents were essential to the relevant standards.

1. 268 Patent—construction. “Frame” in Claim 1 meant a mux data frame. Its meaning was context-dependent, and the claim’s technical purpose required the overhead bits to remain identifiable to the receiver. “Specifies” required the selected parameter to identify which frames contained overhead bits and which did not. The reference to n_max was a non-limiting reference sign, or in any event imposed no meaningful limitation.
2. 268 Patent—validity. Alabama addressed the problem of excessive overhead at low data rates, but its proposal enlarged the mux data frames and retained overhead at the beginning of each frame. It did not teach that some mux data frames contained overhead and others did not, as specified by a parameter. The defendants’ obviousness case impermissibly depended on hindsight and confusion between different processing levels. The 268 Patent was valid.
3. The court rejected the alternative attacks advanced during closing submissions. They had not been properly pleaded or addressed in the evidence, and allowing them would have been unfair. The conditional amendment therefore did not arise, although it would have been allowed if necessary.
4. 268 Patent—infringement. The claim was not limited to the preferred truncation embodiment. In VDSL2, parameter G_p specified which mux data frames contained overhead octets where G_p was less than T_p. The patent was essential to the relevant standards and had been infringed.
5. 430 Patent—construction and infringement. “Normalising” did not require CRC anomalies to be calculated and reported consistently for every connection. The claim contained no such limitation. If valid, the patent would have been infringed because it was essential to the relevant standards.
6. 430 Patent—validity. The skilled team included persons concerned with DSL error reporting and management, including an ISP perspective. The skilled person was deemed to read and assimilate ADSL2/02 properly, while focusing on the error-reporting aspects relevant to the patent. ADSL2/02 made clear that PERp had become variable, while SES reporting continued to use a fixed threshold. It required no invention to perceive the resulting inconsistency. Since the claimed normalisation was accepted to be obvious once the problem was perceived, the 430 Patent was obvious and invalid.