VoIP/bandwith/switch question

SweetbabyJ

Contributor
Veteran XX
I know some of you guys are involved/informed in this area, thought I would throw something out there. What companies make or will be making the newer switches that are able to recognize different codecs (from different providers, VoIP guys and verizon, cingular, etc)? I am aware of Nortel and Lucent, any smaller guys out there that might benefit?



This is from an article I read on it...

"My problem in launching right now is that I need to make sure I set up a network that will be ready for the future. The best VoIP switches now recognize the codec others are using and will use that codec (which will happen in one year) instead of a predetermined one. Incredibly, the switches are just now being updated with enough DSP power to do so. This will happen over the next six to 12 months. There are plenty of companies in VoIP where their switches are obsolete already and must be replaced. In fact, that would be the majority. (I hate to say it, but Lucent (LU: news, chart, profile) and Nortel (NT: news, chart, profile) in North America will be the most successful, by far, in updating and coming out with the right technology, while many others will go bankrupt. The Chinese will be the lowest price with great technology, but with virtually no customer support in the U.S.)
 
um SBJ where id you get that article www.idonthaveafuckingclueabouthenetworkindustry.com ?

Codecs are a standard. Different VoIP implmentations use different codecs for voice quality which is factored into bandwidth,delay and jitter conerns.

I really would not worry about which codec in the next year will be the 'standard' (there will not be one)

Am I answering your question? I'm kinda confused as to what you want to know. Also Cisco (albeit small, plays in VoIP as well)
 
I've heard good things about these guys.

http://www.sysmaster.com/

They only support G.723.1, G.729A/B/AB, G.726 and GSM. I'm not sure which bit rates for G.723.1 they support, I suspect it's 6.3.

G.723.1 (R6.3) and G.729 are the most commonly used CODEC's, at least from my experience excluding wireless carriers.

I'm not really sure what you're really trying to do but CODEC's are a small portion. Are you going to be doing H.323, SIP, MGCP or MEGACO? Does your implementation require RAS signaling? What interoperability requirements do you have? Will you need support for CAS, Q.931 or SS7 on the trunk side? There's a lot more to VoIP than CODEC's. Good luck.
 
Fubar said:
um SBJ where id you get that article www.idonthaveafuckingclueabouthenetworkindustry.com ?

Codecs are a standard. Different VoIP implmentations use different codecs for voice quality which is factored into bandwidth,delay and jitter conerns.

I really would not worry about which codec in the next year will be the 'standard' (there will not be one)

Am I answering your question? I'm kinda confused as to what you want to know. Also Cisco (albeit small, plays in VoIP as well)


I'm not worried about what the standard codecs will be, I'm wondering what companies will benefit from the move to a new technology that accomodates both cell and VoIP. There will be strong demand for this. Does that make sense?

lol this is the article I read about it.


SAN DIEGO (MarketWatch) -- The best thing about Dan Borislow when he ran Tel-Save (now Talk America) in the 1990s, at the time an upstart provider of cheap long-distance service, was he'd answer any and all questions, even if they it was confrontational -- as mine usually were.


And often those answers weren't short or t sugar-coated. He was great copy. Then he quit the company, sold his stock, pocketed a lot of money and galloped into the sunset to become a racehorse owner.

But, horse racing has its limits for an entrepreneur, which can only mean one thing: He's back!

I had heard through the grapevine that Borislow started a new company in the voice-over-Internet (VoIP) business, so I asked him: Is it true you started a new business?

Simple question, right? This is his e-mailed answer -- and trust me, even with the technical nitty-gritty it's a great read, especially in light of eBay's (EBAY: news, chart, profile) Skype acquisition:

"I have been working on a new company for awhile, but have only hired people over the last four months to work with me. I have another idea that will shake things up again if I can get the right vendors to work with me.

"VoIP really has not been ready for primetime, where you can operate a company and make money at this -- until now. With that said, you would have to operate your company to perfection to be successful. There are a number of things that are getting better very quickly, but there are some things that need addressing for the good of our whole telecom system in the U.S. The price of everything involved with VoIP and Wi-Fi is dropping dramatically, with the performance of these products increasing inversely to their price. That is what holds all the promise.

"Here is the problem: We have way too little bandwidth on our cellular and international networks. As a result, they use codecs, which are voice compression techniques using complicated algorithms. That in itself is not the issue, because you still can maintain super voice quality using these codecs. The issue is when the different cellular carriers and VoIP guys use different codecs. Every time you use a different codec, you add delay, because it takes the switch and its DSP time to convert analog to digital. If the carriers are using different codecs, it must be converted once again. Cingular and Verizon, for instance, use different codecs, which is why you will notice a difference of voice quality when you call from a Cingular phone to Cingular phone as opposed to Cingular to Verizon.

"It is getting worse because the more cellular subs there are, the algorithms get more complex and the delay gets worse. VoIP uses different codecs than cell right now. But the huge thing that is happening is that cell companies are about to be the largest buyers of VoIP switches and gateways -- and then the equipment manufacturers will be forced to supply these codes with these switches.

"My problem in launching right now is that I need to make sure I set up a network that will be ready for the future. The best VoIP switches now recognize the codec others are using and will use that codec (which will happen in one year) instead of a predetermined one. Incredibly, the switches are just now being updated with enough DSP power to do so. This will happen over the next six to 12 months. There are plenty of companies in VoIP where their switches are obsolete already and must be replaced. In fact, that would be the majority. (I hate to say it, but Lucent (LU: news, chart, profile) and Nortel (NT: news, chart, profile) in North America will be the most successful, by far, in updating and coming out with the right technology, while many others will go bankrupt. The Chinese will be the lowest price with great technology, but with virtually no customer support in the U.S.)

"At my old company, we had the best voice quality of any network in the country, with the best billing and sign-up capability. I have always believed that besides having a marketing gimmick, you need a product differentiator to be successful over the long term. Tel-Save signed up over 4 million customers on America Online (TWX: news, chart, profile) , and still profits to this day from that deal. That deal was worth in profits close to $1 billion.

"Call me crazy, outspoken -- whatever -- but if you were to collect my public statements and my intense desire to get out of that business when I did, it would seem to make sense today. I was competing against a ghost with Worldcom and CLECs (competitive local-exchange carriers) that were willing to lose tons of money; the dumbest and/or the most crooked telecom and Internet analysts; the worst partner you could have in AOL and, very honestly, a difficult media that was misled by the aforementioned people.

"The proof is that all these people were either arrested, fired, fined or something not pleasant except for the media and Tel-Save. Even after losing the top visionary in telecom, Tel-Save still survives as Talk America (TALK: news, chart, profile) . The reason is that it had the systems and personnel in place , I had bought back over $400 million in debt before I left and there was a very profitable business that was built on my ideas.

"The bottom line is that I was right then and exited at the perfect time. I will do it again (except for the exiting part!) if I can put together a couple of more pieces that are technically feasible to do in the short-term. I believe that I can build something much greater and longer-lasting and that I will do so not only though better ideas, but also because I went from being blue collar to white collar almost overnight and was ill-prepared to deal with these people from AOL, investment houses and the believing media -- and at the same time compete with companies that would lose all of their money to the point of bankruptcy.

"I was right to exit when I did. I was wrong in setting myself up and making public to my shareholders before I exited promises given to myself by an unnamed company, who indicated they were going to buy Tel-Save. I should have known better not to trust what I was hearing. I feel like Steve Jobs in the sense that I have matured and I've learned to make sure I control my own destiny, not to listen to so called experts (analysts), because I know what is best for my company. I will now handle PR better and instead of writing off criticism from the media (and let it get under my skin), I will understand their point of view and rationally explain mine to them.

"Vonage and others need to hire telecom guys to figure out the technology for the future and use much more creative marketing schemes to lower their acquisition costs. My ideas, if executed with precision, will dominate the industry."

So what exactly will this new business be?

"Without giving you my groundbreaking ideas," Borislow says, "the company will most likely be called Talk4free. Just like Tel-Save took existing rates in the industry down about 50% overnight, this will be even better for the consumer. It will be wireless, but I do not like calling it VoIP-like. The quality of conversation will be excellent and the features will be unmatched."

So it'll be a free VoIP? He'll only hint that it could be free phone service in Wi-Fi enabled areas.

How will he make money on something that's free? "If I told you how I make money," he said, sounding like classic Borislow, "then I would have my competitors an opportunity to beat me to the finish line, and I don't like second place."

Should be, in short, quite a race
 
Also, CODEC negotiation should be facilitated by the H.225 stream, if H.323 is used. I believe all the major protocols have some sort of CODEC negotiation built into them.
 
Polarbear said:
I've heard good things about these guys.

http://www.sysmaster.com/

They only support G.723.1, G.729A/B/AB, G.726 and GSM. I'm not sure which bit rates for G.723.1 they support, I suspect it's 6.3.

G.723.1 (R6.3) and G.729 are the most commonly used CODEC's, at least from my experience excluding wireless carriers.

I'm not really sure what you're really trying to do but CODEC's are a small portion. Are you going to be doing H.323, SIP, MGCP or MEGACO? Does your implementation require RAS signaling? What interoperability requirements do you have? Will you need support for CAS, Q.931 or SS7 on the trunk side? There's a lot more to VoIP than CODEC's. Good luck.

I tried to googletranslate this paragraph, and I my computer BSOD'ed :shrug:
 
Usually when you sign up VoIP, the provider either give you a router/box or you have to buy one. My guess is whomever has the cheapest box (with codec built in) with reasonable quality will win in the end. I paid about $150 for a year for my VoIP, it comes with a router box, cordless base and 2 cordless phones. They aren't making much profit since that covers a whole year and I'm getting freebie hardware.

I do get some delays in my conversation but I am not sure it's due to my ISP or it's due to my VoIP provider (not enough bandwith, bad compression/decompression or what not).

If I were you, just pick a good VoIP and let them worry about codec, box, etc.
 
You know skype let you call from computer to computer for free (reminded me of roger wilco, teamspeak, etc. back in tribes days). If you pay a bit of money, skype will let you call someone's phone number at a minimal charge. Skype is getting quite popular so I am not sure if regular telco will be the biggest player in the future, skype's customer base is international. Whatever codec skype uses may have influence.

I know you want to buy stock ro something based on this but based on the info we have, it's hard to say, industry may not swing towards telco but skype as I know more and more people around me are using skype and VoIP phones. We only kept our 1 land line for 911 purposes, that's it. We have no caller ID, call waiting or any feature, it's just a dumb phone for us to make 911 calls and receive calls. We use skype, VoIP phone and cell phones a lot more. Our friends are also signing up to skype and VoIP providers gradually. I can call them on skype or through our internet phone as it is, who cares about Telco if it's not for 911 :)
 
The article you posted talks about VoIP and voice communications from a carrier level. The problem he's referring to is the multiple transcodings that occur over a single conversation.

So you start with normal PSTN, you use a VoIP provider that compresses it using G.729 until they pass it off to a carrier and it's decompressed back to G.711, then it's passed off to a wireless carrier who compresses it to GSM. Every time you convert the stream you experience loss and it's exponential.

So what he's referring to is the ability to bypass the multiple transcodings by compressing the RTP stream once instead of multiple times by compressing it to be appropriate to what the other party uses. That way if it does have to be transcoded again for some bizzare reason the overall number of transcodings is reduced.

This is a problem for anyone who's deploying VoIP. Fortunately two transcodings don't degrade the signal so bad that it's completely unusable so they pretty much just do it. If you're considering deploying VoIP on your LAN for management purposes then use G.711 and you don't have to worry about this because G.711 doesn't compress the signal.
 
Last edited:
Blitzkrieg said:
You know skype let you call from computer to computer for free (reminded me of roger wilco, teamspeak, etc. back in tribes days). If you pay a bit of money, skype will let you call someone's phone number at a minimal charge. Skype is getting quite popular so I am not sure if regular telco will be the biggest player in the future, skype's customer base is international. Whatever codec skype uses may have influence.

I know you want to buy stock ro something based on this but based on the info we have, it's hard to say, industry may not swing towards telco but skype as I know more and more people around me are using skype and VoIP phones. We only kept our 1 land line for 911 purposes, that's it. We have no caller ID, call waiting or any feature, it's just a dumb phone for us to make 911 calls and receive calls. We use skype, VoIP phone and cell phones a lot more. Our friends are also signing up to skype and VoIP providers gradually. I can call them on skype or through our internet phone as it is, who cares about Telco if it's not for 911 :)

eBay bought Skype so it will be interesting to see if Skype survives what's coming.
 
Polarbear said:
The article you posted talks about VoIP and voice communications from a carrier level. The problem he's referring to is the multiple transcodings that occur over a single conversation.

So you start with normal PSTN, you use a VoIP provider that compresses it using G.729 until they pass it off to a carrier and it's decompressed back to G.711, then it's passed off to a wireless carrier who compresses it to GSM. Every time you convert the stream you experience loss and it's exponential.

So what he's referring to is the ability to bypass the multiple transcodings by compressing the RTP stream once instead of multiple times by compressing it to be appropriate to what the other party uses. That way if it does have to be transcoded again for some bizzare reason the overall number of transcodings is reduced.

This is a problem for anyone who's deploying VoIP. Fortunately two transcodings don't degrade the signal so bad that it's completely unusable so they pretty much just do it. If you're considering deploying VoIP on your LAN for management purposes then use G.711 and you don't have to worry about this because G.711 doesn't compress the signal.

The article points out the problems of different codecs that are getting worse exponentially with more cell subs. He's looking to setup a network for the future, all i'm asking is if you guys have an opinion about who will be behind that network, what companies will benefit from this demand? Lucent and Nortel?
 
SweetbabyJ said:
The article points out the problems of different codecs that are getting worse exponentially with more cell subs. He's looking to setup a network for the future, all i'm asking is if you guys have an opinion about who will be behind that network, what companies will benefit from this demand? Lucent and Nortel?

There won't be one network. It's an interoperability issue that has to be worked out between the different carriers. He's trying to setup one VoIP network that addresses this issues with multiple carriers. I don't think anyone can predict who it will affect and who will come out on top.
 
Polarbear said:
There won't be one network. It's an interoperability issue that has to be worked out between the different carriers. He's trying to setup one VoIP network that addresses this issues with multiple carriers. I don't think anyone can predict who it will affect and who will come out on top.

I'm not saying it will be one network, I guess I should have asked -who will be behind the technology to build a network *such as* the one he plans to build.

To address the problem of different codecs that will be getting exponentially worse, *someone* will be creating new technology (the switches I believe?) to accomodate everyone. If I am understanding what you're saying, you aren't sure who would likely be behind this?
 
SweetbabyJ said:
I'm not saying it will be one network, I guess I should have asked -who will be behind the technology to build a network *such as* the one he plans to build.

To address the problem of different codecs that will be getting exponentially worse, *someone* will be creating new technology (the switches I believe?) to accomodate everyone. If I am understanding what you're saying, you aren't sure who would likely be behind this?

Yes, new technology will likely be created to help facilitate this. It's usually dictated by a standards body like the ITU. Then companies like Avaya, Nortel, Cisco and a ton of other less known but large telecom hardware manufacturers will implement the new standard, providing that it fits their overall business plan, and thus the new technology will be available. But the carriers would have to agree upon and start implementing the new technology.

Of course this doesn't always happen. SS7 was supposed to be deployed throughout the telecom network years ago and there's still some DMS100 switches out there that have no SS7 links. It's something that's done voluntarily by the telecom players and money is usually the deciding factor.

I don't know who the biggest manufacturer of Class 5 switches is (I think it's Nortel) and there's a TON of carrier grade switch manufacturers out there. You've got the old Nortel DMS100 and DMS150s, a bunch of old AT&T and Lucent stuff and the mix goes on from there. One company that I would watch if I were you is Nortel. I hope that helps.
 
DocHolliday said:
That was a joke, our system here sucks and the few people that I know that know about VoIP tell me how much Avaya sucks.

In many respects yes. Avaya sucks ass hard. Their VoIP implementation is only partially standardized and it's mixed with proprietary crap. The S800's are nice though, but only in a homogenous environment with a private network.

:wave: Doc
 
Last edited:
Polarbear said:
Yes, new technology will likely be created to help facilitate this. It's usually dictated by a standards body like the ITU. Then companies like Avaya, Nortel, Cisco and a ton of other less known but large telecom hardware manufacturers will implement the new standard, providing that it fits their overall business plan, and thus the new technology will be available. But the carriers would have to agree upon and start implementing the new technology.

Of course this doesn't always happen. SS7 was supposed to be deployed throughout the telecom network years ago and there's still some DMS100 switches out there that have no SS7 links. It's something that's done voluntarily by the telecom players and money is usually the deciding factor.

I don't know who the biggest manufacturer of Class 5 switches is (I think it's Nortel) and there's a TON of carrier grade switch manufacturers out there. You've got the old Nortel DMS100 and DMS150s, a bunch of old AT&T and Lucent stuff and the mix goes on from there. One company that I would watch if I were you is Nortel. I hope that helps.

thanks, that helps. was looking for a smaller co. to potentially invest in but I don't know this industry well and imagine most of the smaller guys went under in the past 5 years.
 
Back
Top