Background
The Commonwealth Scientific and Industrial Research Organisation (CSIRO), an Australian government research agency, held U.S. Patent No. 5,487,069 covering a technique for reducing multipath interference in wireless radio communications — a technology that was incorporated into the IEEE 802.11a wireless networking standard. Because virtually every Wi-Fi device must implement 802.11a to comply with the standard, CSIRO’s patent was effectively a standard-essential patent (SEP). CSIRO pursued infringement claims against Cisco Systems and other major Wi-Fi device manufacturers.
The district court awarded CSIRO approximately $16.2 million in damages against Cisco. Cisco appealed, arguing that the damages award was methodologically flawed in two key ways: (1) the damages model should have been required to start from the smallest saleable patent-practicing unit (SSPPU) — the smallest component in which the patented feature is practiced — rather than the whole Wi-Fi product or chipset; and (2) the royalty rate should have accounted for the fact that the patent was standard-essential and that CSIRO had made FRAND commitments to the IEEE standard-setting organization.
The Court’s Holding
Judge Taranto, writing for the Federal Circuit, addressed both issues. On the SSPPU argument, the court rejected Cisco’s proposed categorical rule that all damages models must begin with the smallest saleable patent-practicing unit. The court held that the SSPPU framework is one permissible methodological approach for ensuring apportionment, but it is not the only approach and is not legally required in all cases. The goal is simply to ensure that damages are apportioned to reflect the economic value attributable to the patented feature — this can be achieved through appropriate royalty base selection, royalty rate adjustment, or a combination of both. A rigid SSPPU requirement would be “untenable” because it might not be the most reliable basis for damages estimation in all circumstances, particularly when the patent covers a key technical advance rather than a minor incremental improvement.
On FRAND and the standard-essential status of the patent, however, the court found error. The district court had not adequately accounted for the fact that CSIRO’s patent was incorporated into the IEEE 802.11a standard and that CSIRO had agreed (through IEEE’s standard-setting process) to license it on FRAND terms. The FRAND commitment is relevant to damages because it reflects a prior agreement by the patent holder to accept reasonable licensing terms — a patent holder cannot use its SEP position as leverage to demand supra-FRAND royalties by exploiting the lock-in that comes with standard adoption. The court vacated the damages award and remanded for a FRAND-adjusted royalty calculation.
Key Takeaways
- There is no categorical requirement that all patent damages models start with the smallest saleable patent-practicing unit — apportionment can be achieved through royalty rate adjustments or other reliable methodologies.
- For standard-essential patents subject to FRAND commitments, the royalty award must reflect the FRAND obligation — a SEP holder cannot use its standardization leverage to extract royalties above what it agreed to accept through FRAND commitments to standards bodies.
- The FRAND commitment affects both the royalty base and royalty rate analysis: the court must consider the value the patent contributes to the standard (and the technology) separate from the value contributed by the standard’s adoption and lock-in.
- The case is a companion to Ericsson v. D-Link (2014) in developing the Federal Circuit’s framework for SEP damages — both cases require that SEP royalties reflect the technology’s intrinsic value, not the hold-up leverage created by standardization.
Why It Matters
CSIRO v. Cisco is important both for its rejection of the mandatory SSPPU starting point and for its treatment of FRAND in damages calculations. On SSPPU, the Federal Circuit made clear that the goal is apportionment — ensuring the patent owner recovers the value of its patented contribution rather than unpatented features — but that mechanical application of any single formula is not required. Damages experts have flexibility to use various methodologies as long as they reliably apportion.
On FRAND, the case reinforces the principle established in Ericsson v. D-Link: SEP holders who have made FRAND commitments cannot ignore those commitments when calculating damages. The court must separate the value of the underlying invention from the lock-in effects of standardization. For companies involved in standards development, the case underscores the strategic importance of FRAND commitments and the lasting effect they have on recoverable royalties. For implementers accused of infringing SEPs, it provides a basis to argue for FRAND-adjusted damages even when the patent holder initially demands higher royalty rates. The combined Ericsson/CSIRO framework remains the primary legal architecture for SEP damages litigation in the U.S. federal courts.