AmberBlog

Marvell Semiconductor had a less than perfect Xmas last year, having had a $1.17 billion judgement against them for infringement of two patents held by Carnegie Mellon University, after a jury trial in Pittsburg.

The award was in relation to two patents that referred to complex algorithms used to convert signals into data, for example when reading the data on hard disc drives. When reading such data, the computer sees a stream of data - but which of this signal is the underlying information and which is noise?

To answer this question, the Italian born American engineer Andrew Viterbi developed the Viterbi algorithm, which is a means of determining the most likely explanation, i.e. the underlying information, that underpins the signal. And then went on to receive a national medal of science for developing this algorithm and co-found Qualcomm, among other companies.

The Viterbi algorithm works in part by forming a structure with branches. Initially it was assumed that the noise in the different branches was independent, but in some circumstances this noise is correlated and this can introduce errors in the final outputs. The two Carnegie Mellon patents claim a means of compensating for this correlated noise.

Claim 4 of US patent 6201839, (Filing year 1998, priority date May 1997, and which we will 'CM1' because it is the earlier patent ) covering: 

A method of determining branch metric values for branches of a trellis for a Viterbi-like detector, comprising:

selecting a branch metric function for each of the branches at a certain time index from a set of signal-dependent branch metric functions; and

applying each of said selected functions to a plurality of signal samples to determine the metric value corresponding to the branch for which the applied branch metric function was selected, wherein each sample corresponds to a different sampling time instant.

Claim 2 of US patent 6438180 (Filing year 1999, same priority date as CM1, hereafter 'CM2'), covers (when combined with its dependent claim 1)

A method of determining branch metric values in a detector, comprising:

 receiving a plurality of time variant signal samples, the signal samples having one of signal-dependent noise, correlated noise, and both signal dependent and correlated noise associated therewith;

selecting a branch metric function at a certain time index; and

applying the selected function to the signal samples to determine the metric values.

wherein the branch metric function is selected from a set of signal-dependent branch metric functions.

In response, Marvel claimed that these patents were anticipated by US patent 6282251 to Seagate, which also discloses the use of a Viterbi algorithm to read the data on a spinning hard disc drive, and where the noise on the different branches in the multi path is correlated, and not random. However Seagate, a competitor to Marvell, was apparently happy to support Carnegie Mellon in this dispute, providing an email from the inventor of it patent noting that the CMU invention went well beyond the current state of the art.

Whether the Seagate patent is valid prior art for two Canargie Mellon patents is a matter for the appeal for this case - and there is some commentary on the web which expresses strong opinions on this. However, we were intrigued to see what AmberScope patent searching could add to this case, starting with the links between Carnegie Mellon patents and Marvell. 

Citation links to Marvell Semiconductor

Carnegie would not had to look far for suggestions that Marvell may need to take a license to their two patents. As a simple example of this, the website Patentbuddy offers the ability to identify and rank companies that have forward citation links to their patents. For the CM1 patent, the leading owner of forward citations was Marvell with 13 forward citations, while CM2 had 24 forward citations from Marvell. Forward citation by themselves are not proof of infringement, but does suggest that the citing company is interested in and has invested in developing similar technology.

It is also worthwhile considering how the Marvel patents are connected to the Carnegie patents. The figure below shows an AmberScope network of the patents connected to the CM2 patent. Patents belonging to Marvell have been marked with green circles using the word 'filter'. 

As you can see, the patents connected by (mostly forward citations lines, which are green as opposed to the blue backward citation connections) to the CM2 patent fall into three main clusters. A cluster means that the patents have interconnections between them, as well as to the CM2 focus patent.

You can also see the Marvel patents are distributed mainly in the bottom right cluster, although some patents are found in the cluster on the left. This cluster of Marvel patents would suggest a number of patents filed for similar inventions by Marvel, which in turn suggests a technology investment by Marvel in this technology area. So an investigator in this area would be well advised to review this patents to see if they described a technology, which if being commercialised by the patent owner, may infringe other patents in this area.

Prior art search - the easy way?

Given the natural interest in this billion dollar duo of patents, we were intrigued to see what the citation data could tell us if there were other patents out there that: 

• Discussed the treatment of correlated noise in a Viterbi algorithm using a time delay
• Had filing dates prior to 1998
• Had not been previously identified as prior art for either of the Carnegie Mellon patents by the USPTO.

We have done this search in three ways:

1) Using line thicknesses  within AmberScope to suggest similar inventions - and 'exploring' onwards from there

The complexity of inventions such as this one puts a big onus on searchers, who may not be fully familiar with the underling technology. Which raises an interesting question - what if AmberScope could analyse the citation information to lead searchers to a potentially relevant, purely in a objective way? Which may never replace the hard and necessary work of drawing final subjective conclusions from reviewing a patent, but could suggest which patents most deserve further analysis.  

In fact this is possible by using line thicknesses in Amberscope networks. Consider the figure below, which shows the network associated with CM1. This network shows forward citations, backward citations, and ghost patents, which are indirectly connected patents that may be similar to the focus patent. 

If we zoom into the area around the focus patent CM1 (which is easy to do using a mouse wheel), it can be seen that the lines joining the patents are of different colours and thicknesses. The colours refer to the citation direction with respect to the patent being reviewed, in this case CM1. Green lines point to forward citations, and blue lines to backward citations. 

The line thickness refers to the predicted similarity of the two patents, which is calculated from other citation relationships in the wider network. In this, the thickest line is to the other Carnegie Mellon patent, CM2. This should not be surprising, as CM2 is a continuation in part patent of CM1.

But how can this be used to find prior art patents? CM1 has a filing date of 1998, and so we will filter out all all patents later than 1998 using the filing year circled below. Only a few patents are left, and all are backward citations. One line is slightly thicker than others, see below.

This patent is US5689532 (filed 1996), filed by Quantum Corporation and now owned by Seagate. This patent does not specifically mention correlation of white noise and does not appear to be a direct  disclosure of CM1 - this is not surprising, as being a known backward citation for CM1, CM1 may not have been granted if the Quantum patent was a direct disclosure.

However the value of the Quantum patent is it too has connections, in this case to some ghost patents to the CM1 patent. The strongest connection (by line thickness) is to US5949831 (filed 1997) filed by IBM and now owned by HGST Netherlands, a subsidiary of Western Digital. 

This discloses a similar technology, namely that "Colored (non-random) noise at the input of a PRML Viterbi sequence detector results in sub-optimal performance." using a matching delay circuit to provide a delayed PR4 Viterbi output signal.

2) Searching using ghost patents

There are various means of using AmberScope effectively. One of the more useful and innovative features of AmberScope is it ability to find ghost patent, which are defined by our system as 'second order' patents (patents connected to patents connected to a focus patent, or 'friends of friends') that have a series of good connections to first order (directly connected) patents. We think they are useful because they are 'new' information, and so possibly can be used an opposing party to request re-examination of a patent. Ghost patents can be easily identified in an AmberScope network as they are faded out and so 'ghost like' in their appearance, as shown in the figure below, and highlighted by a red circle.

Although there are two Carnegie Mellon patents, we can start our analysis with CM1. So what the ghost patents for this patent?

In the example below, all of these ghost patents have have given a relevancy score of 0 (purely to colour them green) and been added into the table below, which has been downloaded and formatted below. For comparison, we have also added the Seagate patent referred to above in this table.

One of these patents is the US5949831 patent identified earlier. As this technology is quite complex, technical specialists in this area may be required to confirm the ultimate relevance of this potential prior art to the two Carnegie Mellon patents.

As you can tell from Table 1, some of these ghost patent appear to disclose some of the key elements of Carnegie Mellon patents, yet have not been cited by the patent examiner. It is of up to the courts to decide whether these other citations are relevant to Carnegie Mellon patents, but this shows how quickly AmberScope can identify potentially relevant prior art.

A further observation from this table is that the two Carnegie Mellon patents do not have particularly high AmberScore values. CM2 has an "AmberScore" (a measure of network connectiveness) of 4.4, which is 4.4 times the average AmberScore for US patents granted in the last 20 years. 4.4 is obviously better than average, but would not ordinarily predict a billion dollar infringement.

Marvell have made a similar point on their website, noting that the 50 cent per chip royalty failed to consider the:

'value attributable to that (patented by Carnegie Mellon) functionality, as opposed to other improvements.....more than 80 additional features were added at the same time that the (patented) feature was added.'

Obviously any such statements released by a litigant in the middle of litigation are likely to reflect the view and interests of the litigant. However this simple analysis would suggest that the Carnegie Mellon patents are not the most important in this 'neighbourhood', either in absolute (4.4 is not that high an AmberScore value) or relative terms (there were more dominant patents in the immediate neighbourhood).

$1.1 billion seems to me to be a very high royalty payable on a duo of patents that are far from dominant in their space.  It will be interesting to see how the appeal plays out.

 3) Using the already identified Seagate patent as the basis of an AmberScope search

Another option for the prior art search could be to start an AmberScope search from the already identifed Seagate patent US6282251. A preliminary search on this patent admittedly failed to identify anything directly relevant, but we did not fully explore this area. This still remains an option for searchers closer to this case to pursue.

 Conclusions

This blog has reviewed a pair of Carnegie Mellon patents that have been litigated against Marvell Semiconductor. AmberScope was used to identify some potential prior art, and provide some perspective on the relative value of these patents. Overall, notwithstanding the complexity of the technology and value of hard disc drives, this analysis has suggested that the billion dollar infringement cannot be supported by the relative dominance of these patents. We await the appeal with interest

One of the challenges in searching patents and patent landscapes is working with crowded patent landscapes, or patents with lots of citation connections. 

For example, consider patent US4799156 (1986), filed by Strategic Processing, and covering an Interactive market management system:

A system for interactive on-line electronic communications and processing of business transactions between a plurality of different types of independent users including at least a plurality of sellers, and a plurality of buyers, as well as financial institutions, and freight service providers. Each user can communicate with the system from remote terminals adapted to access communication links and the system may include remote terminals adapted for storage of a remote data base. The system includes a data base which contains user information. The data base is accessed via a validation procedure to permit business transactions in an interactive on-line mode between users during interactive business transaction sessions wherein one party to the transaction is specifically selected by the other party. The system permits concurrent interactive business transaction sessions between different users.

This appears to be an early patent for e-commerce, for example describing how Ebay appears to work. Espacenet lists 1417 forward citations  and 10 backward citations, while the USPTO lists 1390 forward citations:

Our AmberScore value, which measures the overall connectiveness of the patent, is 137, or 137 times as connected as the average granted US patent. This was enough to earn it a place in our recent top 10 patents of the 1980s blog. This patent appears to be one of the key early patents for e-commerce, and so a review of its forward citations should provide a good understanding of developments in e-commerce.

However, from a practical viewpoint, this does raise a couple of important questions:

• How can you practically review a list of 1400 odd forward citations?
• In particular, which of these 1400 connected patents are most likely to be most worth looking at?

Readers of this blog should be familiar with AmberScope, which is a way of visualising the citation network around a patent via an intuitive interactive graphical display. However up into very recently, AmberScope too would have struggled with such a highly connected patent network, and would have ended up showing more information that could be usefully ascertained. This would also have slowed down the display of the network, and some computers might have struggled to display such a complex network.

But not any more. A very recent improvement to AmberScope has changed that the way that 'complex' (more than 100 connected and ghost patents) patent maps are shown.

By 'top patents', we refer to the highest ranked patents according to AmberScore, which is the algorithm we have developed to measure the network influence of a patent. We believe that patents with higher value of AmberScore are more likely to disclose important inventions. This is why the our filtering mechanism focuses on the top ranked patents. Similarly, other highly ranked patents are likely to be connected to these highly ranked patents. 

You will note that there are 100 patents shown (see the red circle) - and AmberScope has done this by automatically selecting a value for the % filter that leads to 100 patents - see the red arrow pointing to the % filter.

AmberScope automatically selects this % range to show the top 100 patents. And if you may want to look at more than the top ranked patents, you can simply adjust this % filter to any other values, as in the figure below where by moving the % filter so that the top 200 or so patents are displayed.

Alternatively you can choose to filter by filing year range - the figure below shows the top 53 patents in the 10 years after this patent was filed.

Returning to the top 100 odd connected patent, all years, there is an easy of reviewing these, namely using the 'Next' button found in the bottle left hand corner of AmberScope

This automatically selects the highest ranked patent in the displayed network that has not yet been read. This happens to be US4674044 to Merrill Lynch (AmberScore value of 113) for an Automated Securities Trading System.

When you look at this summary box, you will note a sub-box reading '484 more':

What this label refers to is that there are 484 further patents connected to this Merrill Lynch patent that are not in the network connected to US4799156  - providing a link to further patents in this area.

And we could keep going. The next highest ranked patent is US4903201 to World Energy Exchange Corporation, for an Automated futures trading exchange. This has an AmberScore value of 110, and is a 'ghost patent', i.e was not directly cited by the examiner for the '156 patent, but is still likely be to relevant (note that the yellow patent highlighted below is slightly transparent in nature, showing it is a ghost patent). Ghost patents are of particular value to people aiming to invalidate a patent, but given that the '156 patent is now expired, this may not be an issue.

Of course this is all a little backward looking. What if we wanted to look at more recent patents, say filing in the most recent decade? In the image below, we have set the filing date filter to be between 2000 and 2010, and adjusted the % filter to show the top 50 patents in this age range. Ebay is an obvious contender is this area, and so we have entered ebay in the word highlighter box (which highlights all patents with this term in the title, owner or user comment box). Among the Ebay patents meeting this criteria (with green circles around them) is US7593866. 

Makes searching a crowded patent landscape very easy, doesn't it? And has avoided all of the potential risks of otherwise filtering these patents using filters based on keywords and patent codes.

Test out the new 'auto-filter' to search crowded patent landscapes at AmberScope.com, and make the most of the free beta trial while it still lasts.

Do you need to search every patent in a patent network?

While in same cases you should, you do not need to in all cases. Instead my experience is that simply reviewing the highest ranked patents using the Next button may find you what you are looking for.

As an analogy, consider a Google search, for say hybrid cars. You might run this search in Google:

But you do need to search through all 113 million results? No, of course not. Instead you will likely search the first page or two of results, and find probably find a page that gives you what you wanted when you started the search.

Using the Next button to search results when using AmberScope is similar to this, i.e. you may not need to search for every patent. 

Unless of course you are doing a freedom to operate search - but Ambercite products are not recommended for freedom to operate searches in any case, unless being to complement more traditional patent searching processes.

 * Currently AmberScope is not listing very recent forward citations as our data is about 14 months out of date. We are working on making our data completely up to date, and so this anomaly is only temporary. 

'Patent landscape' is a term often used by a range of people in the IP profession. But what exactly is a patent landscape in practice? And why are they increasingly being produced?

Why are patent landscapes so useful?

Patent landscapes can provide a very unique and valuable perspective on a technology and its commercial interests.

Patent applicants file patents in order to protect their innovation. Governments grant patents for patentable inventions in order to a) encourage investment in innovation and b) to encourage publication of new ideas. This second effect should not be underestimated, as even the most secretive of companies can file thousands of patents, providing access to their thinking that that otherwise would be impossible to obtain. Further helping the quality of patent data is the fact that patents can cost significant amounts to file, meaning that applicants only file patents for what they perceive to be potentially commercially valuable and worthwhile inventions.

According to the latest data available from WIPO, 1.9 million patents were filed in 2010, to add to the more than 70 million patents already filed. Patents are often made available from government run databases in a highly systematic manner, making them ideal for future analysis. Patent landscapes can provide an technology and commercial overview simply unavailable from any other source.

What exactly is a patent landscape?

WIPO has defined 'patent landscape' as meaning 'an overview of patenting activity in a field of technology, and published a range of patent landscape reports on their website

By reviewing these and other published patent landscape reports, the most common approaches start to come out:

1) The objective of the patent landscape is defined. This normally has some commercial, public policy or marketing objective, and in some cases can include some non-patent perspectives on the objective

2) The objective is used to define a relevant technology

3) The definition of the technology is used to create a patent search aimed at finding patents relevant to the technology. The search query can include a combination of keywords and patent classification (IPC or USPTO). Patents can be limited to those filed in certain countries, or in recent years. Identified patents can be enhanced by adding patents linked by forward or backward citations to the identified patents

4) In some landscapes patents are combined into patent families, for example when the same patent is filed in a number of different countries. In other landscapes, the individual patents are kept separate.

5) A process can be used to remove 'false positives', i.e. patents picked up in the search terms that are not relevant to the project objective. This can be automated (for example based on looking for certain keywords or patent codes that are clearly irrelevant to the objective) or manual, where a reviewer manually removes irrelevant patents. Alternatively a combination of automated and manual processes can be used.

6) An attempt is made to standardise the owners of patents. For example, patents owned by TOYOTA USA may be combined with patents owned by TOYOTA JAPAN on the basis that they belong to the same overall company.

7) Patents are group in a meaningful way. For example a patent landscape report in the area of renewable energy patents might cluster patents into by the type of renewable energy. In some cases, this taxonomy of patent groups can include sub-groups, for example the aforementioned set of renewable energy patents would include a grouping of solar energy patents, and then could include sub-groups for different types of solar energy production. Clustering can be done in an automatic fashion, for example by keywords or patent codes. Or patents can be manually assigned to clustered, or again a combination of automated and manual processes can be used. Patents can also be grouped by owner, country of origin, age, and status, and in some cases grouped using a combination of parameters, i.e. "solar cells patents filed by Japanese companies". Arguably this grouping and sub-grouping of patents is one of the most important parts of patent landscaping, as this help uncover patterns in the patent data.

8) Trends graphs can be produced to look at major filing trends, either for the overall number of patents, or just for certain groupings.

Figure 1): Filing trends for hybrid car patents. © Griffith Hack 1999. Griffith Hack images and examples will be used in this many examples in this blog instead of images from other and equally worthy authors simply to avoid any copyright issues.

9) The leading patent owners in a technology area can be identified and listed. This helps to identify who are likely to be most important companies in the technology. An alternative and also worthy approach is to identify the leading patent owners in specific groupings of patents

Figure 2:  Leadings owners in patent landscape study of Alzheimer's patents, including a breakdown into patent groupings. © Griffith Hack 2012.

10) The leading inventors in an area, or for an applicant, may be identified

11) The leading sources (where the patents have come from) and destination (in which countries the patents have been filed in) may be identified

Figure 3: Leading country of origin for carbon trading related patents. ©Griffith Hack 2012.

12) An attempt can be made to rank the patents in terms of quality. Rankings based on forward citation count and family size are the most commonly used, but different analysts can use a variety of ranking techniques. Network Patent Analysis (NPA) ranks patents based on the influence they have on a network of related patents, and has been shown to be powerful predictor of patent quality.

13) Details can be provided of some particularly interesting patents

14) Citation analysis can be used to show relationships between patents or patent owners

Figure 4:  Forward and backward citations for Motorola (now Google owned) patent US6,246,862. ©Ambercite 2012.

15) An attempt to be made to find 'white space' in the patent landscape, i.e. areas in the patent landscape where the relative number of patent filings is low.

Figure 5: NPA white space analysis. © Ambercite 2012.

Patent landscaping outputs

Outputs of patent landscapes can be in several forms

a) Reports

b) Graphs

c) Spreadsheets listing patent details, which including patent quality rankings and details of patent groupings

d) Patent landscape images, in which similar patents are clustered together . These come in two forms. The most common form of patent landscape maps cluster patents by keywords. At Ambercite, we group patents using a unique process of clustering patents based on patent citation linkages as we believe that this is more precise and inclusive than clustering patents based on keywords.

Ambercite is very proud to be associated with the latest Griffith Hack NPA white paper Clearing the fog: Patenting trends for the treatment of Alzheimer's disease, which was released today. In this white paper Griffith Hack, working very closely with and applying the Network Patent Analysis (NPA) process developed by Ambercite, analyses over 48,000 patents to fiilter, cluster and rank these patents. Two separate NPA maps accompany the white paper, one NPA map showing an cluster focused patent landscape map, and one NPA map showing a time scale patent landscape map.

Clearing the fog is also the best publically available demonstration yet of the powerful ability of NPA to precisely cluster patents with a precision unavailable with keyword or IPC patent code clustering. While we have seen this precise clustering for the majority of the confidential NPA client studies we have delivered, this degree of clustering is stronger than in our previous two NPA white papers on hybrid car and smartphone patents.

Clearing the fog also also demonstrates several other features of NPA:

• the power of associative searching (page 5)
• the value of a NPA time scale map (page 15, and available as a separate download)
• the concept of foundation patents (page 16)
• the ability of NPA to identify what could important future patents (page 17)
• and even some natural limitations of NPA (page 21)

As well as a detailed discussion of the leading patents, patent owners and inventors in the area of Alzheimer's disease, an increasing important disease which may impact many of our elderly and the people that care for them.

Interested in learning more, or how NPA can be applied into your business? Come back to us, and we can share more about the NPA process and deliverables. 

Network Patent Analysis™ (NPA™) is the sophisticated analysis and mapping of patent citation data for the purposes of determining the leading patents, patent applicants and technology trends in any area of technology. In this white paper, NPA is applied to the patents wars currently underway between major smartphone companies such as Apple, Nokia and Motorola to answer such questions as ‘who has the leading patents?’, and ‘which areas are being litigated the most?’.

Read the full reports here.