EFF Patent Busting - Prior Art Needed for VOIP

JumperCable writes "The Electronic Frontier Foundation is seeking to bust an overly broad patent by a company called Acceris. Acceris claims patents on processes that implement voice-over-Internet protocol (VoIP) using analog phones as endpoints. These patents cover telephone calls over the Internet. Specifically, the claims describe a system that connects two parties where the receiving party does not need to have a computer or an Internet connection, but the call is routed in part through the Internet or any other 'public computer network'. The calls must also be 'full duplex', meaning that both parties can listen and talk at the same time, like in an ordinary phone call. To bust these overly broad claims, we need 'prior art' — any publication, article, patent or other public writing that describes the same or similar ideas being implemented before September 20, 1995."

Comments

Phone patches for radio?

[MichaelSmith]

Specifically, the claims describe a system that connects two parties where the receiving party does not need to have a computer or an Internet connection, but the call is routed in part through the Internet or any other 'public computer network'.

In CB radio, and possibly Amateur (Ham) radio you could have a phone patch device which would interface between the radio transciever and the phone system. With two such gadgets you could bridge a gap in the PSTN. Not really legal with amateur radio as you were not supposed to compete with commercial services.

I am sure that emergency services used phone patches on their VHF radios, though. Some documentation on that might be of some use.

TFA talks about it being full duplex. The impression I have is that this system would have used one frequency and a VOX to switch between transmit and recieve. It is possible there were true full duplex systems though.

Re:Phone patches for radio?

[NfoCipher]

>In CB radio, and possibly Amateur (Ham)
You've got that reversed.

>Not really legal with amateur radio as you were not supposed to compete with commercial services.
Autopatch has been and still is "legal".

Re:Phone patches for radio?

[Andy_R]

Not over the internet, or using intetnet protocol, so it's not VOIP.

(note to mods: I know I've posted this 3 times in reply to different people, but I maintain it's not redundant until people actually grok the concept and stop posting/modding up non VOIP references.)

Here's the oldest reference I got

[mangu]

The way it was described in the blurb, I guess the oldest implementation is mentioned here.

Maybe they have the answer themselves

[vivaoporto]

Maybe EFF already has the answer, depending on how long AT&T is routing all phone calls through NSA network. They would even kill two birds with one shot, the subpoena to obligate AT&T to disclose the info could come from the patent suit. It's a win-win! What could possibly go wrong?

VOIP Prior Art

[azrider]

Not sure if it was patented, but in the 70's when I worked for IBM, all office extensions worldwide went through the "tie-line". This was a linkup that used the massive IBM internal global network to make calls, i.e. I call Tokyo from LA and the call never touches the PSTN apparatus. Indeed, it never left the building on anything other than data lines. The phones at the desks were plain old analog WE2500 sets.

Re:VOIP Prior Art

[Andy_R]

Sorry, if the IBM system never touches PSTN as you describe, then this fails part 4 of the EFF's list of features the prior art needs to have:

From the EFF site: CRITICAL FEATURES OF PRIOR ART NEEDED:

      1. The system must have the ability to connect an audio telephone call from a calling party to a receiving party.
      2. The telephone call must be "full duplex," meaning that both parties must be able to talk and listen at the same time. For example, regular telephone calls usually are full duplex, whereas walkie-talkie conversations in which a person cannot receive transmissions from others while he or she is transmitting generally are not.
      3. An ordinary telephone and telephone line are the only equipment the receiving party needs to have. The receiving party does not need to have a computer or an Internet connection to receive the call.
      4. The transmission of the call is routed in part through a "public computer network" and in part through the PSTN. This implies that the transmission must cross at least one gateway between the "public computer network" and the PSTN. The Internet is one example of a "public computer network," but the patent does not define what else would qualify as a "public computer network."

Additional Features:

      1. The caller must only have to dial the destination number and no additional phone numbers

Prior art should NOT be the problem.

[mrchaotica]

This is ridiculous. All this patent covers is bridging between the Internet and POTS networks. It shouldn't need "prior art" to be struck down, it should be struck down merely because it's fucking obvious! I mean, it'd be one thing if it were a patent on one particular clever method of connecting the two networks, but the idea in general should not have been patentable in the first place.

Re:Prior art should NOT be the problem.

[pla]

All this patent covers is bridging between the Internet and POTS networks. It shouldn't need "prior art" to be struck down, it should be struck down merely because it's fucking obvious!

I don't think it does count as that obvious. If you remember the earliest days of free internet telephony, the biggest limitation (aside from the annoying lag) came from needing both parties to have a computer with an always-on connection (or risk missing calls).

Companies like Vonage exist to make a free service un-free solely because they act as a POTS bridge. Think about that. People will pay for something free (well, "free" presuming you would have intenet access anyway) because that one little "fucking obvious" step counts as such a massive leap forward in functionality.



The USPTO has made some phenomenally bad calls in the past, but I don't know if I can really disagree with this one.

Re:Prior art should NOT be the problem.

[billcopc]

The "fucking obvious" problem is that you need a number to call to/from, and that number is not an IP address. The technical aspect of VoIP companies is trivial, just convert the sound to digital and ship it off to its destination... but the way things are today, you still need a POTS at the receiving end if they don't have VoIP. That's not technically difficult, it just requires money to rent the landlines from Ma Bell. VoIP "carriers" oversell landlines much like ISPs oversell bandwidth or modem banks, because not everyone uses everything all the time.

If all your friends and relatives, employers, partners, and local businesses were on VoIP, we wouldn't need the gateway because the PSTN would be obsolete. We're not there yet, and I don't think we'll ever be, not in my lifetime, because computers are unreliable. Even the standalone VoIP gateways are just embedded computers running outsourced software, they crap out every once in a while. The day they crap out on an emergency call is the day it will all come crashing down. The phone companies may be the axis of evil, but they usually provide quite dependable service, something the internet still hasn't mastered.

Re:Prior art should NOT be the problem.

[jambarama]

The problem isn't the *obvious* issue. I mean, it wasn't obvious to me in 1995, or most other people I'd wager. The problem is the scope of the patent. No one should be able to patent "processes that implement voice-over-Internet protocol (VoIP) using analog phones as endpoints." It is way to broad. Acceris should hold a patent on A SINGLE process to implement VoIP. You shouldn't be able to patent an end result, just the specific way you used to get there. Patents like this make clean room reverse engineering, work arounds, and competing methods all illegal without the patent holder's permission.

Re:Prior art should NOT be the problem.

[arth1]

The problem isn't the *obvious* issue. I mean, it wasn't obvious to me in 1995, or most other people I'd wager.

Back in '94, I was talking over speakfreely to an overseas friend on my Indy when my mother dropped by, and asked what I was doing. I told her, and she thought that would be horribly expensive since I was talking to someone on the other side of the planet. When I told her that it used the internet connection, so I only paid for the internet connection (mind you, a 128 kbps BRI was expensive enough back then), she asked whether I could hook up a regular phone to it. I had to explain that it was technically possible, but required the cooperation of phone companies on both sides of the internet line, so with the sluggishness big corporations operate with, it wouldn't happen any time soon.
Mind, she was an old lady who never even figured out how to use half the functions of her remote control or microwave oven, and yet to her it was THE obvious use.

Artisoft LANtastic could do this

[scsirob]

How about this link: http://www.pcmag.com/article2/0,1895,1161458,00.as p

It describes a voice adapter for Artisoft LANTastic in 1990. I used to operate a LANtastic network but didn't use the voice adapters. However, it seems to fit the 'prior art(isoft)' requirement ;-)

Re:Artisoft LANtastic could do this

[gravis777]

Interesting, but from the way this product is described, its LAN use only, which means that it does not connect to a public network, and it does not seem to connect at some point to the public phone network, which means it canot be used in this case

Re:Vocaltek?

[Andy_R]

Yes, the linked article says the EFF are specifically looking for proof that VocalTec or Net2Phone were doing this before 20th September 1995.

Graham Article

[Rob_Warwick]

I'm not sure if this qualifies, since the article wasn't written until 2005, but Paul Graham mentions in one of this articles that a friend of his wrote some VoIP software in 1994. The article is available online.

In 1994 my friend Koling wanted to talk to his girlfriend in Taiwan, and to save long-distance bills he wrote some software that would convert sound to data packets that could be sent over the Internet. We weren't sure at the time whether this was a proper use of the Internet, which was still then a quasi-government entity. What he was doing is now called VoIP, and it is a huge and rapidly growing business.

Re:Graham Article

[mavenguy]

But did it involve using the PSTN at both ends? Just computer to computer via is acknowledged by the patent as prior art; the central point of novelty is the use of plain telephone sets at both ends communicating with each respective CO; that is regular duplex telephone traffic routed to a local service that converts both ends to/from a connection over IP to a similar remote service that converts back to an appropriate duplex analog telephone traffic to the remote party's analog telephone.

Break Stupid Laws

[essence]

Not sure if this would work, it would probably just end up in people getting sued bigtime, but what if there was a 'class action' of sorts, where a whole heap, and I mean heap, of people/companies used patented ideas, and basically told the patent office and the patent holders to get fucked. It would take co-ordination, but done on a mass scale, the point could be made, and the patent system reformed.

Mod parent anything but insightful, it's funny

[vivaoporto]

Come on, people don't recognize humor when they see it anymore? Next time I'll be telling that Microsoft has given up zune and will pay people to use it and I will be modded informative. Hmmm, wait!

Re:ISDN

[Icarus1919]

Yes, if wikipedia says it, it's so. Case closed, indeed.

I have an old Russian book, dated 1986

[WetCat]

Samoilenko S.I "Seti EVM" (Computer Networks), Moscow, Nauka, 1986

which describes Adaptive Communication (connecting voice phones using packet switching).
This book also referencing
Bellamy J.C. Digital Telephony. John Wiley and Sons, 1982

May be something can be found in that book too?

Re:ISDN

[synoniem]

What is the difference between a connection over a X25 network and "the internet"? Especially in the early days of internet X25 networks were used a lot.

Re:Vocaltek?

[JonathanR]

The company I worked for (Automotive parts manufacturing) between Sept '95 and Jan '97 had a system where interstate (non-local) calls could be routed through their leased data lines. There was a dialling prefix for each endpoint node. The data lines were ordinarily used for warehouse inventory/stock control operations (I think it was a VAX/VMS system, so I'm not sure what networking protocols were used for these data links). This was introduced halfway through my period of employment there, and given the conservative nature of the company, I'd be surprised if this was a bleeding edge installation at the time. Obviously it would have been developed well before this implementation.

Simon Hackett's Etherphone?

[Russ Nelson]

I wonder if Simon Hackett's Etherphone qualifies? He was running voice calls over raw Ethernet packets back in 1992. He wrote up a white paper which was distributed at Interop that year.

Re:Vocaltek?

[Spamalope]

We did that at Leasing Associates from roughly 1993 until early 2006 using Datarace (brand) equipment - voice, compressed and full duplex. Certainly voice as a sideband over whatever leased line you've got was very common before 1995. All of the mux manufacturers had equipment to do it. The datarace equipment could route TCP/IP, but the equipment encapsulated each type of traffic and sent it to the other mux via a proprietary protocol to the other mux.

There were ISDN router boxes touted around that time as being able to route voice and data at the same time, using small packets, compression, and QOS to keep the voice from breaking up. The setups I remember were analog->voip->analog though, and were used to tie corporate phone systems.

It sound like this is a patent on existing packet switched voice tech, but specifying which devices the endpoints would be and what the transport protocol would be. Private PC->voip-PC was common, and private analog->voip->analog was common. There were regulatory barriers to doing PC->voip->analog AND tying into the PSTN (public phone system). Privately our 1993 system allow a branch office to press a one button extension on their phone, get a dial tone from the corp. office phone system, and make any call they wanted to. It just wasn't done with TCP/IP because other protocols are much more efficient.

Telecom 95 in Genève

[dybdahl]

I attended the Telecom 95 exhibition in Genève, and I still remember how the news went around, that the finnish telephone backbone would be expanded using IP-capable equipment, to carry both internet traffic and telephone calls. This seemed very logical at that time, for those who knew about TCP/IP. I cannot believe that such huge investments in using the IP protocol for telephone traffic was made, unless the decision makers had seen internet telephony work. This means, that there is prior art somewhere.

I suggest that you look into the PR messages released at the Telecom 95 exhibition, and then do some research on those that cover telephony over TCP/IP.

The obvious prior art is...

[freebase]

The existing Public Switched Telephone network.

I've not read the patent, but if the claim is really as broad as indicated, it would seem to include the PSTN currently used for 'analog' calls.

The PSTN, by definition a Public Network, is made up of analog access lines connection analog 'terminals' - your phones - to what's known as a Class 5 switch. Class 5 switches are connected together at what's known as a Tandem, providing connectivity between all the users within an area. Access to the long distance network is via a connection to a Class 4 switch, usually at the tandem, but not always. Class 4 switches are interconnected (internetworked??) with other switches, and eventually a sufficient network is formed that allows you to call anyone with a phone.

The Switches (Class 5, Class 4, etc) used in this network are very much computers, and have been for quite some time http://en.wikipedia.org/wiki/5ESS_switch.

The analog to digital conversion used to be done in the CO itself, and sometimes still is, but usually it's done at the Digital Loop Carrier (DLC) closest to the customer http://en.wikipedia.org/wiki/Digital_loop_carrier.

This network even has its own routing and control protocol, SS7 http://en.wikipedia.org/wiki/SS7.

Plainly, the only thing really new about VoIP is that it abtracts the physical transport and allows the control plane traffic to be transmitted on the same path as the bearer plane traffic.

Only solution

[jmorris42]

> Haven't phone companies been running phone calls over digital networks for ages?

Yes they have, and in a sane world that would in itself have ended the discussion at the USPTO. Since the first telco stuff was crude circuit switched equipment a better example would be ATM, which also easily predates the patent. But apparently the USPTO and the courts are still allowing a fresh patent for doing ordinary old things by simply adding "over the Internet" to them. We seriously need a law of one single paragraph:

"No patent may be issued or upheld if the only thing unique about it is that it extending an existing practice to the Internet. This is one of the designed purposes for the Internet; using something for it's designed purpose is NOT original or difficult for one skilled in the art so knock it off you idiots. This law is intended both as an order to the USPTO and binding guidance for the Judicary."

Found one!

[DigitAl56K]

Please google "1994 gsm over ip"

http://goliath.ecnext.com/coms2/gi_0199-721578/ip- access-and-RigNet-deliver.html

M2 PRESSWIRE-24 February 2004-ip.access: ip.access and RigNet deliver GSM Abis over IP via satellite; ip.access and RigNet partner for implementation of GSM-over-IP-over-satellite solution; Successful trial paves way for delivery of GSM services to remote locations(C)1994-2004 M2 COMMUNICATIONS LTD

Also looks interesting:
http://kbs.cs.tu-berlin.de/~jutta/toast.html
http://kbs.cs.tu-berlin.de/~jutta/gsm/toast-igp.ht ml

CDMA (IS-95) carried voice over IP

[claykarmel]

I believe IS-95, the first publicly used version of CDMA, which was in public use in 1995, carries voice packets over a TCP/IP network from the phone to the mobile switch. From there, the full duplex phone call terminates on any phone on the PSTN.

The question is whether the Sprint or Verizon IS-95 infrastructure constitutes a 'public network'. I would think so.

Wikipedia includes a lot of detail about IS-95, as do books on CDMA available on Amazon, so presumably Qualcomm does not mind publication of high level characterizations of it. I also sat through classes in CDMA at UCSD which described IS-95 in glowing detail. So I have good reason to believe none of this is confidential. EIA/TIA/IS-95 and IS-99 and IS-707 are published specifications available from Global Engineering.

I learned about this TCP/IP network in 1996 while developing 'data devices' to run on the IS-99 (data) overlay of IS-95. In order to present a TCP/IP socket to a handset application (which could terminate anywhere on the web), we had to run an additional TCP/IP stack. That is, our application formed a PPP connection to TCP, wrapped in IP, then PPP again, which was wrapped in the lower stack TCP and the lower stack IP. The lower stack terminated at the mobile switch (enhanced to handle IS-99), with L2TP or PPTP connection to an IP router. The upper stack terminated on a web server. It seemed like an insanely complex link, but it worked surprisingly well because of the highly tuned TCP/IP stack running on the Qualcomm chipset. (I think this connection was later branded as "QuickNetConnect".)

That is, the lower stack wasn't there for data. It was there, I believe, for Voice (Over IP) services in IS-95.