Monday, July 28, 2014

iBiquity Vs. Patent Trolls

The legal complaint is as dry as most any other. It is rife with what is pursuant to whom, proper venues, infringements, hereins and obligations to indemnify. But it gets exciting for one single solitary sentence.  "Had the patent examiner known or been made aware of these prior art references, the claims would not have been allowed and the [5506866]patent would not have issued." Aside from the vicarious joy of punting a patent troll around the DC beltway, how is this legal spat important?

Ok so it still sounds dry but what iBiquity  is saying as politely as they can is that the patent examiner is an ignorant clod. Instead of arguing just that Wyncomm are malevolent twats, iBiquity is essentially arguing that our patent system is broken and stupid. While both of these things are true, I think it's the second assertion which has legs. Technology patents are arcane, complex, and nearly impossible for a layperson to understand. I'm not saying the patent examiner is lazy, or dim, or didn't make a sincere effort. The problem is that compared to a radio engineer, mathematician, or C++ programmer a lay patent examiner doesn't have the proverbial snowballs chance in hell of understanding all the technical nuances of a complex patent.
More here So let us examine the material in question: U.S. Patent Numbers 5506866, 5642379 and 5475691.

5506866 - Filing Date 11- 15-1993
It originally belonged to AT&T but was used as loan collateral and was held by both Patent Business Development LLC and Clearwater Innovations LLC in recent years. The patent covers a narrow use of side-channel communication which can carry simultaneous voice and data transmission. Here they argue that they invented a technique to simultaneously transmit voice and data while also transmitting additional information separate and apart from the voice and data information. In webcasting we might call that third stream metadata. In broadcast radio we might call it RDS. They are describing transmitting data while broadcasting in CDMA. Ridiculous. Telegraph signal multiplexing dates to at least the 1870s, telephone multiplexing to at least the work of George Owen Squier in 1910. Prior art is an understatement. At the same time confusing this for IBOC (NRSC-5) is patently ridiculous.

5642379- Filing Date 06-14-1993
This patent describes a technique for modulating orthogonal signals with one or more analog or digital signals. Don't be intimidated by the word orthogonal. Even stereo LPs exhibit orthogonal signal. It just means having two overlapping but uncorrelated signals. Both TDMA and OFDM can be described as orthagonal. Any simultaneous transmission of two signals could be considered prior art. In this patent they describe voice and data, but it lists a large number of virtually identical patents: 4280020, 4313197, 3824347, 3875339 and others. This tries to solve the non-existent problem of FFT bandwidth. It does so with a common bandpass filter. Please just stop.

5475691 - Filing Date 11- 15-1993
Voice activated date rate change in simultaneous voice and data transmission. This technique is said to improve the transmission quality of the voice signal by dynamically adjusting bits per symbol. This is already used in everything from Ethernet LAN to HDTV.  Trying to increase this number is essentially the history of all IT hardware between your modem and your ISP.  This one is beneath our contempt.