TP-Link Systems Inc & Anor v Huawei Technologies Co Ltd

MR. JUSTICE MEADE:

1. This is my judgment on the Defendant's (“ Huawei ”) application for a declaration that its adjustable licence offer is RAND or, alternatively, determination of the RAND terms for an adjustable licence between the parties. I will refer to this as the " Interim Licence Application ". The Interim Licence Application is brought in the context of an implementer-led claim by which the Claimants, who I will refer to as " TP-Link ", seek a final RAND licence to Huawei's portfolio of Wi-Fi 6 and previous generation Wi-Fi patents that are essential or are claimed as essential to standards organised by the Institute of Electrical and Electronics Engineers (the " IEEE ") and I will refer to this as the " Huawei Portfolio ".

2. Huawei has committed to license the Huawei Portfolio on RAND terms. Occasionally, I may refer to “FRAND” rather than “RAND” below. That is just a function of the different terms used by various other SSOs and in previous judgments in this area of the law and does not reflect any point of substance. TP-Link is a manufacturer of consumer products that implement Wi-Fi 6 and previous Wi-Fi standards. I am told that no other patents are sought so be licensed and there is no cross-licence sought.

3. Mr. Lykiardopoulos, King's Counsel, appeared for TP-Link, leading Mr. Lunt. Mr. Chacksfield, King's Counsel, appeared for Huawei, leading Mr. Cronan. I am grateful for the help of all the advisers on both sides.

4. At the hearing, the parties relied on the following evidence. For Huawei, three witness statements from William Warne (a partner at Huawei's solicitors) of 9th December 2025 and 8th and 11th January 2026, and for TP-Link, four witness statements from Annsley Ward (a partner at TP-Link's solicitors) of 18th and 31st December 2025 and 9th and 12th January 2026.

5. By way of background, Huawei commenced proceedings in China in late 2023 and late 2024, together with patent infringement proceedings before the UPC and the German national courts commenced in August 2025. TP-Link then commenced these proceeding in this Court in September 2025 and following the commencement of these proceedings, Huawei submitted to the jurisdiction of this Court and consented to TP-Link's claim that this Court should settle the RAND terms of a global licence between the parties, with the licence to run through to a currently confidential end date some years away.

6. Huawei has also agreed an interim licence regime should be put in place pending the UK RAND trial and that when signed by TP-Link, Huawei will dismiss, withdraw, and discontinue all actions against TP-Link around the world. At the hearing, TP-Link raised the fact that Huawei had not yet dismissed or withdrawn the various global proceedings, but that is simply because TP-Link has not yet signed the licence. Huawei's position that it will withdraw the proceedings when and only when the interim licence is actually signed is reasonable.

7. The central issue for this hearing was the amount that TP-Link should pay under the interim licence, which the parties are agreed that I should decide. The litigation situation

8. Although of course ideally the parties would settle their whole dispute, for the moment they feel they have to litigate. Given that there is litigation, however, the situation that has been arrived at is a relatively good one. Huawei is guaranteed to be paid a RAND amount decided by a neutral tribunal with an interim payment in the meantime and TP-Link is spared from any litigation pressure which might otherwise lead to hold up. The parties and other courts internationally will not have to waste time and money on pointless infringement or validity litigation. There is no international friction between courts and no risk of offending comity either by anti-suit injunctions or non-consensual interim licence decisions. The parties' agreement also means that there is no need to expedite the RAND trial. As it happens, the parties have agreed the UK as the forum to determine the RAND terms, but there is no reason why that had to be so and no compulsion. The parties' positions

9. For the purpose of these proceedings, in November 2025, Huawei voluntarily provided early disclosure of some of its Wi-Fi-specific out-licence agreements. TP-Link has applied for wider disclosure of licences, in particular including licences where the Wi-Fi technology is licensed with other technology, such as cellular or video codecs, which, says TP-Link, would provide further useful data points and reveal, TP-Link says, the real value that Huawei attributes to the Huawei Portfolio. This wider disclosure application is not for determining on this Interim Licence Application.

10. Huawei's position is that the RAND royalties should be at a standard or rack per unit rate for Wi-Fi 6 units and a lower per unit rate for pre-Wi-Fi 6 units. Huawei says those rates are supported by its disclosed licences to which I have just referred. Applying these rates to IDC data going back to 2008 (when TP-Link first started selling) along with interest at US Prime + 1%, Huawei calculates the RAND royalty payment as a lump sum for the past (2008 up to 1st January 2025) and then a running royalty after that date for the period from 1st January 2025 to the end date of the RAND licence. In the alternative, Huawei calculates the payment for the future period as a lump sum, but its primary position is that it should be a running royalty for that future period.

11. At the hearing I made clear that I thought a running royalty for the future as part of the interim licence would be excessively complicated and Huawei did not press it, especially once the position about white goods (see below) was clarified. I will say no more about this and proceed on the basis that the interim award will be a lump sum for the past and for the future. This in no way precludes Huawei for arguing for a running royalty for the final RAND licence and the principle that the patentee can choose between possible RAND formats for its licences will no doubt be invoked by Huawei at trial.

12. TP-Link's position is that the final RAND licence should provide for a lump sum royalty of US$12 million. This is based on a written offer of April 2025 using a top-down or profits available approach. Huawei says that offer lacked necessary information and logic and is fallacious. Huawei's total final RAND payment for which it contends is a significant multiple of the US$12 million. TP-Link agrees that the US$12 million should be kept by Huawei whatever the final RAND amount since the final RAND amount can be no less on ether side's case. Accordingly, TP-Link agrees that US$12 million of the interim amount paid should be non-refundable.

13. As to the interim adjustable licence, TP-Link has unpacked some, but not all, of the Huawei early disclosure licences by reference to 2023 IDC sales data to give ranges of what it says are weighted average per unit rates that are significantly lower than Huawei's, both for pre-Wi-Fi 6 units and for Wi-Fi 6 units, which it then applies to TP-Link sales covered by the period of the interim licence, which TP-Link contends should be 2018 (reflecting limitation periods) through to the second quarter of 2027 (the likely date of the result of the RAND trial), and uses that to determine the additional amount to be paid above the US$12 million. I will call this " TP-Link's Unpacking Analysis ".

14. TP-Link's Unpacking Analysis gives a number for the interim licence quite a bit higher than the US$12 million, but a lot lower than what Huawei seeks. TP-Link's Unpacking Analysis was developed in the evidence of Ms. Ward. It is complex on any view. Because of the confidentiality issues, no expert was involved and TP-Link itself was not involved. Ms. Ward's evidence is answered by that of Mr. Warne and again no expert input was provided. Mr. Warne's criticisms are also complex.

15. TP-Link opposes Huawei's inclusion of interest on past sales for the purpose of the interim licence. In contrast, Huawei's position on the interim adjustable licence is a mid-point approach, whereby the interim licence amount splits the difference between Huawei's rack rate comparable-based numbers with the lump sum or running royalty on the one hand and either TP-Link's US$12 million approach or the numbers from TP-Link's Unpacking Analysis on the other.

16. The upshot is that the difference between the parties at the extreme ends as to what should be paid by way of interim lump sum pending adjustment at the final RAND trial is a factor of about four or five. As I said, TP-Link have put forward US$12 million as the non-refundable amount and that was the basis the parties were working on for some time. Indeed, it was specifically pleaded by Huawei until Huawei, for the purposes of this hearing, said that the non-refundable amount should be TP-Link's lowest realistic case on TP-Link’s Unpacking Analysis and that is significantly more than US$12 million.

17. A couple of matters narrowed at the start of the hearing. The first was to do with an issue referred to as the white label goods issue, i.e. where TP-Link manufactures and sells goods that make use of the Huawei Portfolio to third parties which then brand and sell those goods on as their own. Although not entirely clear before the hearing, by the start of the hearing, TP-Link accepted that the interim licence should cover white label goods and Huawei agreed. TP-Link was unable to provide a quantitative estimate for the purposes of interim calculations, but was working towards providing Huawei with such a figure and said that it was confident it was a small amount.

18. This position unfolded today, i.e. the date of judgment, in a way which I will refer to again at the end of my reasons. This all tied into the non-refundable payment to be made. In their pleadings, as I have said, the parties agreed to the non-refundable amount being US$12 million, and as I have already mentioned above, Huawei sought to increase this amount in reliance on TP-Link’s Unpacking Analysis. Following brief submissions at the start of the hearing, I held that for this Court to consider increasing the non-refundable amount that the parties had, by agreement, been working to, would effectively be the same as summarily dismissing any plea by TP-Link that would keep the final award at trial down to US$12 million. To justifiably increase the non-refundable amount, Huawei would have had to show that TP-Link had no real prospect of succeeding in keeping the final payment down to US$12 million. Given that it was clear since the first round of evidence before Christmas that TP-Link was relying on US$12 million and Huawei had agreed to it as the non-refundable amount, I considered it to be unfair and bad case management to defer any argument on a change in the non-refundable amount to a further hearing, which Huawei suggested in case a further hearing was anyway needed for the white label goods point, and Huawei's change of position had come too late to allow TP-Link proper time to prepare to meet it at the hearing of the Interim Licence Application.

19. In any event, Huawei will get the full appropriate amount at the end of the final RAND trial, so there is minimal prejudice to not allowing it to argue for an increased non-refundable amount now. Accordingly, I held that the non-refundable amount should stay at the agreed US$12 million. This comes with one caveat. As I said at the hearing, Huawei has liberty to apply to adjust this amount upwards if the aforementioned figures for white label goods are significant. That should be a simple pro rata increase to cater for something not previously disclosed by TP-Link and does not involve any change of basic approach, so it would not mean that the Court would have to assess whether or not TP-Link's underlying logical position on the minimum amount is arguable.

20. There was also a detailed dispute between the parties to do with withholding tax. By the end of the hearing, this seemed to wither away, and the parties seemed to be in agreement as a matter of principle: while US withholding tax would have to be deducted by TP-Link when making a payment, it is Huawei's liability. So if Huawei ends up having to repay TP-Link any of the interim amount, Huawei will have to repay gross and then reclaim from the relevant tax authority. As I made clear during the hearing I am not going to rule on the detailed tax treatment or wording of the documents which are matters for the parties and primarily for the paying party.

21. Consequently, the issues that I have to resolve in order to determine the amount to be paid on an interim basis to hold the ring pending a final RAND trial are as follows: (1) whether to adopt the mid-point approach, which is Huawei's position, or TP-Link’s Unpacking Analysis, which is TP-Link's position; (2) if I adopt the mid-point approach, what figure to use as the lower bound, a US$12 million-based approach, as TP-Link argues in that scenario, or the result from TP-Link’s Unpacking Analysis, as Huawei argues; (3) whether the interim licence should run to when the judgment from the final RAND trial will likely be known (that is the second quarter of 2027), or until the end of the licence term which will be significantly later; (4) whether the interim payment should go back to 2008 (when TP-Link sales started) or 2018 (the latter being tied to limitation periods); (5) whether the interim payment should include interest, as Huawei contends, or whether matters to do with interest should await the final RAND trial as TP-Link contends; and (6) if there should be interest, what the rate should be. Applicable legal principles

22. There is a growing body of case-law in this jurisdiction as to when to grant an interim licence declaration and what the terms should be. In the present case, as I have explained above, the parties are agreed an interim licence should be granted and the dispute concerns what the terms should be. On determining an interim payment amount the parties acknowledged that the cases have so far relied on the mid-point between the latest offers from each side. See for example Panasonic v. Xiaomi [2024] EWCA Civ 1143 at [100] and Lenovo v. Ericsson [2025] EWCA Civ 182 at [156]. However, the case law has also made clear that this does not exclude the possibility of another methodology. See the following passage from Samsung v. ZTE [2025] EWCA Civ 1383 : "6. The Court considered that the grant of the declaration would serve a useful purpose because it would be persuasive, albeit not coercive. It would not be contrary to comity because, if it led to an interim licence being agreed, it would relieve the foreign courts of burdensome and wasteful litigation; and, if not, it would be up to the foreign courts to make their own assessments of the parties' conduct. As to what terms for an interim licence would be FRAND, in both Panasonic v Xiaomi and Lenovo v Ericsson the Court essentially took a midpoint between the parties' most recent royalty rate offers and applied that rate to the anticipated period of the interim licence. (This does not exclude the possibility of another solution in an appropriate case.)"

23. In its written submissions and at the hearing, Huawei submitted that the use of an alternative methodology to calculating the interim payment, i.e. alternative to the mid-point approach, should not be lightly done but may arise, for example, where a party completely fails to substantiate an arguable position or its position is wrong as a matter of established principle or law. I do not understand TP-Link materially to disagree with this as a matter of principle, but it simply says that the mid-point approach is not prescriptive and has been adopted where there is a lack of an alternative at an interim stage and bearing in mind the need to avoid a mini-trial or to delve into unreasonable detail. See for example Acer, Asustek , Hisense v. Nokia [2025] EWHC 3331 (Pat) at [450]-[452], where Mellor J, faced with a choice between an interim payment based on the mid-point between Nokia's published rates and the codec pool licence rates on the one hand, as contended for by Nokia, or simply applying a 50% uplift to the codec pool licence rates on the other, as contended for by Acer/Asustek, considered that the only option open to him was to adopt the rate at the mid-point between the rate proposed by each side, as he was not in a position to decide whether Nokia or the Claimants were right or, for that matter, which of them might be closer to the RAND rate.

24. I think the case law makes very clear that the question of what terms are RAND for an interim licence is quite different to the question of what terms are RAND for a final licence, with the setting of the former being a more limited task to hold the ring pending determination of the final licence. See the following passage from Alcatel/Nokia v. Amazon [2025] EWCA Civ 43: "75. As Amazon submit, Panasonic v Xiaomi demonstrates that the judge’s reasoning was based on an incorrect premise: Amazon’s claim to an interim licence would not lead to two RAND trials of the same scope. The RAND trial which the judge directed to be tried in October 2025 will involve a full Investigation of all issues relating to RAND. It will undoubtedly be a substantial trial, and it is presently estimated that it will require 20 days of court time. By contrast, Amazon’s application for an interim licence will not require the court to determine most of the issues which will arise at the RAND trial. It will simply require the Court to determine (1) whether Amazon are entitled to an interim licence and (2) if so, what terms are appropriate. As can be seen from Panasonic v Xiaomi , the question of what terms are (F)RAND for an interim licence is quite different to the question of what terms are (F)RAND for a final licence, and determining such terms is a much more limited task. As Amazon submit, this is because the interim licence is only designed to hold the ring pending determination of the terms of the final licence, and the payments made pursuant to it will be adjusted to the extent necessary in consequence of the determination of the terms of the final licence."

25. TP-Link, however, does criticise the mid-point approach on the basis that a rigid application of split the difference between parties' offers would lead to incentives for SEP holders to publish higher rack rates and no incentive for them to make lower offers. TP-Link bases this submission on what was said by the UPC Local Division Mannheim in its 30th September 2025 decision in Amazon v. InterDigital : "48. Of particular relevance in the present case is that, pursuant to Article 20 EPCU, the UPC must give priority to the enforcement of European law. This is only possible if the proceedings concerning a standard-essential patent are conducted in compliance with the EU antitrust law applicable in the present case within the meaning of Articles 101 and 102 TFEU and any questions requiring clarification can be referred to the ECJ pursuant to Article 267 TFEU. These questions also include the question of whether the enforcement of prohibition rights under SEP is in conformity with EU antitrust law. This addresses a central area of regulation of patent law applicable to the EU internal market. One of the questions to be decided is whether, in negotiations for a FRAND license and in calculating its amount and the factors to be applied for this purpose, which are used in a comparison with third-party licenses in order to establish comparability, the SEP proprietor applies criteria that comply with antitrust law. Conversely, the Declaration may result in the SEP proprietor being de facto forced to accept an offer (at least for the time being) that is at the lower end of the FRAND corridor or, depending on the amount of the implementer's offer, even outside it. This question must also be answered in the context of EU antitrust law. If an EU court, in this case the UPC, were prevented from conducting this review, this could result in courts not bound by EU law making determinations on (F)RAND licenses that cannot be legally upheld in the European single market and may even be contrary to public policy. This is particularly to be feared in the present case because the granting of an interim license is not preceded by an examination of whether the competing offers are FRAND-compliant or not. If, therefore, the appropriate interim license rate is simply set at the midpoint between the competing offers, this may result in the determination falling outside the corridor of EU law. It is true that the actual FRAND compliance can be determined subsequently. However, the UK courts do not aim to do this, but consider it desirable for the parties to reach a settlement under the pressure of the interim license. In this case, the amount specified in the interim license may become a reference point in further negotiations that is contrary to EU antitrust law. Incidentally, it also seems doubtful whether this approach will encourage the parties to negotiate more rationally in the future. Rather, it could lead to the exact opposite, with both parties making even more difficult-to-bridge maximum demands in order to have the most favorable starting point possible for further negotiations when overcoming their differences."

26. There are two concerns here: one, which the UPC mentioned second, is that an interim approach based on the mid-point may encourage the parties to make offers that are unreasonably low or high. The other, which the UPC mentioned first, is that rates set on the halfway basis may be non-FRAND and hence contrary to competition law.

27. As to the first, the UK courts’ now considerable experience of these cases show that in real FRAND negotiations the parties, with no thought to interim awards, start very far apart anyway. Having an interim licence regime using the mid-point approach therefore cannot reasonably be expected to make the existing position worse. Furthermore, although parties in real FRAND negotiations start far apart, experience again shows that they usually – indeed nearly always – provide a logic for their offers, not least because the Court of Justice's decision in Huawei v. ZTE is considered to require them to do so. If a party made an artificially low or high offer in negotiations with no support at all, then I think the UK courts could and quite probably would ignore it when making an interim award. In any case, I doubt if parties who are genuinely negotiating global terms will shape their approach just to affect the interim licence that the UK or other courts may set. They will have bigger considerations in mind. Still less I think will licensors set rack rates for their whole licensing business with the UK interim licence approach in mind.

28. As to the UPC's second concern, a RAND interim licence, as the UK courts have seen it, is until a final RAND trial (or a RAND agreement if that supervenes). The eventual final result determined by the court is therefore by definition RAND and the interim licence is just a down-payment on it. So it is inherently impossible for there to be a final non-RAND result set by the UK courts. The UPC's concern is more indirect than that, however. It is that the parties might negotiate a final RAND settlement between the interim award and the later RAND trial, which would or might not be RAND because it would be affected by the interim award.

29. In the passage from Amazon v. InterDigital above, the UPC refers to the settlement being under the pressure of the interim award. I respectfully disagree. The interim award will be RAND in the sense of what willing parties would agree as an interim payment. That is the basis on which it is set. It secures the position that the patentee is not kept out of its money at a time when the implementer is using the patentee's technology and, usually, as in this case, has been doing so for some time. The jurisprudence of the UK courts, of the Court of Justice, and indeed of the UPC has consistently been that it is unfair for the implementer to defer payment indefinitely to the prejudice of the patentee. In addition, an interim award at the mid-point cannot anchor the parties to any particular analysis or decision of what final RAND must be. It specifically avoids making a judgment on who is right or wrong in their analysis; see further below.

30. As to the suggestion that the parties are put under pressure by an interim award at the mid-point, I do not understand what the pressure might consist of and I note that neither party before me has said there would be any such pressure. They are free to continue to negotiate and, given an interim licence, those negotiations will take place in the absence of any risk of an injunction leading to hold up.

31. Further, and in any event, an interim payment, referred to in that context as "security", is also expressly envisaged by the Court of Justice in Huawei v. ZTE . So one can safely infer that the Court of Justice did not think it was un-FRAND to have one and I cannot see how a court which assesses what it ought to be in the light of that Court of Justice case law can be achieving a result which is not compatible with competition law.

32. I am also, I have to say, unclear what alternative the UPC has in mind. Of course, in its Amazon v. InterDigital decisions, it has been hostile to the interim licence declaration approach as a whole, but if there is to be any interim award or security, then some approach to quantification has to be adopted, unless one is to say that the patentee must in all circumstances await a final decision before it gets any money at all, which I think is inconsistent with all the case law and not FRAND.

33. In my view, the mid-point approach is a fair, simple, and robust one which can be done quickly and cheaply and without a mini-trial. It is essential that the interim procedure should be capable of being done quickly, since otherwise the parties might as well wait for the final decision, which would frustrate the whole point of the exercise. I emphasise that the mid-point approach is not the only possibility, as I have said already above in dealing with the case law. The court should be astute to be flexible because the mid-point approach may not be just, fair, and rational in all cases. That could particularly arise if and to the extent that one side or the other is putting forward a position which is arbitrary, artificial, completely unsupported, or obviously wrong.

34. In the present case, I am not in a position to say that the TP-Link top-down approach is obviously wrong and nor am I able to say that Huawei's argument for its rack rates is obviously wrong. But for reasons I will explain, I think that TP-Link's position on interest and the date to which the interim award should go back are obviously wrong and/or not materially supported and I am going to set the interim award accordingly.

35. I think there is another reason for preferring the mid-point approach. It is recognised that there are dangers in a court at an interim stage getting into the relative merits of an argument which it has concluded is at least arguable. As Butcher J said in Magomedov v. TPG [2023] EWHC 2655 (Comm) at [84]: "Moreover, I have also taken the approach that, in cases in which I have found there to be a good arguable case against a Defendant, it is best not then to give any much more detailed analysis of the strengths and weaknesses of the case made against that Defendant, unless necessary to deal with the position of another Defendant. Any such analysis would be likely to be overtaken by what will emerge during the course of the case and/or, as it was put by Knox J in In Re a Company 005009 of 1987 [1988] 4 BCC 424, be such as 'merely [to] embarrass the judge who will have to determine the question at the trial'. That was said in the context of a strike out, and the relevant considerations are not identical, but it nevertheless appears to me to be apt in the present context, and to be, as was said by Lloyd J in Bank of America Trust v Morris (22 October 1988), 'wise guidance'." Standard of proof

36. Huawei sought to make a black and white argument about the applicable standard of proof by analogy to the approach which applies when deciding interim declarations, in particular the high degree of assurance standard versus the ordinary civil standard. It does not matter to my decision in the end, but I do not accept this. Setting the non-refundable amount should be to the summary judgment standard for reasons given above. Otherwise, I think the court should apply a flexible approach to the evidence so as to ensure the interim licence is fair in amount and justly set. Usually, that will mean taking out of account arguments which are completely unsupported or contrary to established principles and by not getting into deciding matters which are arguable either way, such as which basic approach to quantification is better. Past sales

37. As to past sales, in Panasonic v. Xiaomi [2024] EWCA Civ 1143 Arnold LJ summarised the following legal principles regarding limitation and interest, drawing from his judgment in InterDigital v. Lenovo [2024] EWCA Civ 743 : "23. This Court held in InterDigital v Lenovo that limitation provisions under national law had no role to play in the determination of what terms were FRAND, and thus royalties should be paid in respect of the whole period during which the implementer has been exploiting the SEP holder’s portfolio. Among the reasons I gave for reaching that conclusion were the following: '187. … An implementer … requires a licence from the first day it implements the relevant standard(s). FRAND terms are the terms that would be agreed between a willing licensor not intent upon hold up and a willing licensee not intent upon hold out. The ETSI Guide and FAQs page make it clear that a willing licensee would not sit back and wait for demands from SEP owners, but would pro-actively contact SEP owners (whose identities can readily be ascertained from ETSI), and would put money aside for the payment of royalties … It follows that, in an ideal world, the parties should be able to agree terms not long after the implementer has started implementing the standard, or at all events before the expiry of six years from that date. Recognising that the world is not ideal, a willing licensor and a willing licensee would begin by negotiating a standstill agreement in order to ensure that the passage of time during the course of negotiations did not affect the substantive terms ultimately agreed. On that basis, the relevant date for the purpose of determining what terms were FRAND would at the latest be the date of first contact between the parties …

188. Furthermore, as the Supreme Court held in UPSC at [105]–[127] …, FRAND terms reflect the value of the SEPs in the portfolio and must be available to any market participant. It follows … that they should not depend on the date on when the licence is entered into. There should be no discrimination in favour of implementers who are slow to take a licence and against implementers who are quick to take a licence. If anything, it should be the other way around.’

24. This Court also held in InterDigital v Lenovo that FRAND terms required the payment of interest by the implementer in respect of past sales in order to reflect the time value of money. In that case the Court upheld the judge’s decision that the appropriate interest rate was 4% compounded quarterly."

38. See also Alcatel at [57] and Ericsson at [23] and [24].

39. In InterDigital v. Lenovo at [198], Arnold LJ said that if a party could establish in evidence that there is a settled industry practice of releasing past sales that could be relevant to what is FRAND. There was some suggestion by Huawei during the hearing that that statement by Arnold LJ has been superseded by what he said on limitation in the decisions cited above. I disagree and I find no support for that contention.

40. I think it is possible that in the right case an implementer may be able to show an industry practice of releasing past sales, justifying not going all the way back to the start of sales, either for the purposes of final RAND determination or an interim award. This would, in my view, however, need to be an accepted practice and not just an event that has happened on some occasions. I am sure it has happened on some occasions that parties have agreed to release past sales and there has been evidence of that in many RAND cases before the UK courts.

41. At the moment, all there is from TP-Link is a bare assertion of industry practice without any detailed support or any evidence. I do not think that is good enough to take into account when settling the interim amount to be paid and I contrast that with TP-Link's assertion of its top-down approach in its April 2025 offer, which is reasoned and supported to at least some extent. This, of course, does not preclude TP-Link from putting forward a better evidential position at trial as to an industry practice but dragging down the interim award by a bare assertion would, in my view, be unprincipled at this stage. How far forward

42. As to how far forward should be covered by the interim licence, the authorities to date have declared interim licence terms calculated to run until the likely date of the final RAND licence determination at trial, rather than a later date such as the end of the final licence that is proposed: see Panasonic at [100], Ericsson at [102] to [104], and Acer at [455(i)]. This has been on the basis that the interim award is to hold the ring and I see no reason to depart from that approach on this application. Interest and escrow

43. The authorities are clear following InterDigital v. Lenovo in particular that a final FRAND award should include truly compensatory interest. Interim awards should reflect that. TP-Link's only basis for opposing interest being included in the interim award in this case was that the Court could award it in the final RAND award so there would be no injustice in leaving it out now. This argument was only weakly made and, in my view, is obviously wrong. If it were accepted it would remove the basis for any interim payment ever being made. In addition, I think there is generally at least some value to the patentee in having the money in its hands, including what will often be a very significant amount of interest. Huawei's evidence on this hearing was that while an interim payment may not be fully recognisable in its account, it does have value and utility as cash at bank available for use. I note that is a less stark position than that which was taken by the patentee in Lenovo v. Ericsson .

44. For these various reasons, as well as in relation to interest, I also made clear at the hearing and maintain that I would reject any argument that the interim payment should be paid into escrow and not given to Huawei. There is genuine value to Huawei in actually having the money.

45. I see no reason not to award interest at the rate supported in evidence by Huawei, which is US Prime + 1%. Mid-point approach or TP-Link’s Unpacking Analysis

46. TP-Link's top-down position is not as well-articulated as might be wished and has some gaps and potential issues. It may well not succeed at trial, but I am not in a position to reject it now. It is arguable.

47. Likewise, Huawei's main position, that it will get a final RAND award based on its rack rates, may also fail at trial, on the basis that the real effective rate which it achieves is lower or often lower, as may or may not be shown by unpacking of the currently disclosed licences, other licences, or otherwise. But I cannot decide that now and conclude that Huawei's position is also arguable.

48. On that basis, and the given principles I have identified above, the mid-point between those two positions is appropriate.

49. TP-Link’s Unpacking Analysis is, in my view, not an appropriate way to get to an interim award. It is not sufficiently reliable because it has no expert input, it is highly complex and its assessment would require a mini-trial, as was amply demonstrated by the complexity of the arguments at the hearing before me. It would, however, be unprincipled to take a mid-point between TP-Link's Unpacking Analysis and the rack rate approach, as Huawei argues, because TP-Link’s Unpacking Analysis is not TP-Link's lowest arguable position. It is an attempt, instead, to get to the "right" answer. To take halfway between that and Huawei's maximum possible recovery pleaded for would be to take a halfway point between two things which are fundamentally different in nature.

50. Either to try to work out which of the parties' main pleaded cases is better or stronger or to assess the details of TP-Link’s Unpacking Analysis would also, in my view, fall into the trap I identified by reference to the Magomedov case above. Once I have identified that the parties' best pleaded positions are arguable, that is good enough to apply the mid-point approach, which is fair and workable. Getting part-way into the merits would only be a hostage to fortune by hindering the free analysis of the position by the judge at trial, who will have all the materials. The amount

51. I asked the parties to calculate the resulting amount on the basis indicated above: A mid-point between the number implied by the $12 million offer and the rack rate approach, adjusting so that it goes back to the start of sales in 2008 and forwards to April 2027, with interest on the basis contended for by Huawei. There was some discussion in the skeletons for the application about allowing for the fact that Huawei will get early receipt for the royalties between now and trial, although the topic was not, as I recall, revisited in oral submissions. Huawei proposed a 10% reduction for future royalties which will, as a result of my decision above, relate only to the relatively short period from now until the second quarter of 2027 and this should also be reflected in the calculation.

52. There will be a Form of Order hearing which will need to finalise the precise amount if not agreed, deal with disclosure and reply RAND pleadings. As to the Form of Order itself, either the operative part of the order or an undertaking or the interim licence in its contractual terms must contain a general provision allowing the Court to require such repayment by Huawei as it considers just following the RAND trial. My preference is for an undertaking paralleling the cross-undertaking which is given in relation to an interim injunction, as it is simple in expression and flexible, but I will hear argument about the point if necessary. I raised this issue at the hearing of the application, but Huawei did not at the time have instructions to give an undertaking.

53. The directions hearing will also need to deal with costs once the parties have had a chance to consider this judgment and it will need to deal with confidentiality. At the moment, I have structured this judgment so that the end date of the licence, the rack rates and the total amount, are all not public, but I will require each of these to be justified as being maintained as confidential by the parties at the Form of Order hearing if they wish to do so.

54. Finally, by way of postscript, I need to deal with the fact that as explained to me by Mr. Cronan, who appears for Huawei today, that TP-Link's solicitors have, this morning, sent a letter which appears to justify TP-Link's understanding that the number of white goods would be relatively small, but also seems to disclose really very large sales of products referred to as mesh products, which are not, it seems, included in the IDC data used to date. I have already said before I gave my reasons that I am not going to get into, today, the question of how these numbers were omitted. It is perfectly possible that it was an innocent mistake on TP-Link's part or just a misunderstanding and I pass no comment on that.

55. It may be that all that needs doing in the light of the new figures is to treat the mesh products, which I think are in the nature of Wi-Fi repeaters, in exactly the same way as all of the router-type products already the subject of the numbers in the case, in which case the calculation that I have asked the parties to do just needs to be done on a larger number of products and no further decision of principle is needed, but the parties will need to consider that and it can also be taken into account at the Form of Order hearing.