Signal Quality

This is something I wrote to the UTStarCom wi-fi phone forum.  You have to sign up for a support account and jump through hoops galore just to get to their forum, so I thought I’d share it openly as a sort of review.

In 10 words or less: Don’t buy it if you don’t have $200 to waste.

It holds promise, but it’s very far from a ‘working’ implementation.

It’s glitzy. It’s glamourous. The physical design of the phone is quite nice. I’m actually okay with the speaker volume (although volume controls for more than just ‘ringtone’ would be handy), but the rest of it, so far, has left me unenthusiastic.Issues so far:

1. WEP doesn’t work. I’ve tried it on numerous different kinds of wireless routers. 64 bit and 128 bit keys. No dice. When WEP is active, the phone simply doesn’t grab an IP and do what it’s supposed to. (while this might not matter to many, I live in a world where my network needs to be WEP to accomodate my WPA-less Sony AIBO that’s connected to my wireless network… that and there are many WEP networks out there that people might want to use).

2. Poor design on keepalive. Right now, the only way for the phone to keep a NAT connection alive is to reregister every X seconds in order to keep data going through the NAT tunnel so it doesn’t time out. This is a bit like thwarting modem timeouts by redialing/connecting/logging on every 60 seconds. SIP’s registration was NOT designed as a NAT keepalive, and to use it for such is a pretty poor hack. There are options requests, NAT pings, and various other methods (subscribe, notify, etc) which actually WERE designed more to keep a network connection alive, and which don’t require a reauthentication handshake every 60 seconds to keep a tunnel open.

3. Overall connection issues. Connecting and reconnecting to a wireless hotspot when going in and out of range is a bit like pulling teeth. UTStarCom advertises the wondrous nature of how their phone can roam from one AP to another without dropping a call, seamlessly and magically. In reality, I have a hard enough time connecting to ONE AP, much less taxing the poor phone’s logic to reconnect.

4. Features disappearing from spec. Downloadable ringtones? Downloadable wallpaper? These things just sort of vanished from the list of possibilities. Will they ever come back? Who can say…

5. Documentation. The documentation is about like looking at an API function reference. It’s a book filled with one or two line descriptions of each function, without examples, without anything detailed enough for the average user to know what logic was in the minds of the developers when they created feature X. While that may be just fine for someone buying a phone from a retailer who’s bundling it with a preconfigured service plan, for anyone else, it’s less than helpful.

6. Did I mention a lack of something as simple as onboard volume controls with granularity? Oh… wait… I did. Perhaps I just felt it needed to be reiterated. The idea that I can’t control the volume of any tones without altering some sort of global ring parameter is just… odd.

7. Logical mystery. I’ve seen the explanation of why the phone can’t be allowed to connect to SIP servers that have their SIP ports at, say, port 80, but as a former TCP developer, I’m somewhat astounded that there aren’t more people who balked at the poor explanation. If I connect to a SIP server at port 80, the connection has to come back to the phone’s port 80 and would hit the webserver? Does this mean I can’t send email from a machine running an email server, as return communications would hit the email server and not the client? Or run a web browser from a machine running a web server? That someone even suggested a user get a new provider because of this ridiculous limitation in the UA is laughably negligent. If I’d seen one of my developers give that line to ANYone, he’d have been fired on the spot. “I’m sorry, but this car only goes forward and reverse. It has no steering wheel. I recommend you only drive on roads that require no turns.”

Seriously…. this seems like it could potentially be a great phone, but either its limitations need to be taken seriously and overcome, or there needs to be more detailed and valid information about its limitations available somewhere so that future buyers can make the informed decision to go with another hardware provider.

I’ll keep mine on the off chance that someday it might actually be worth using (after all, since it was research to see if we wanted to bundle it with our services, it’s tax-deductible), but for now, it sits in the corner with the other hardware that didn’t live up to its promises.

I was recently asked to give some comments on what I thought of the above proposed regulation (which is essentially ‘USF for VoIP’). I thought I’d share my comments here publicly.

1. What burdens does the proposed rule place on small businesses?

The burden on a small business is tremendous. Small businesses are currently offering low-cost solutions for consumers by maintaining a low overhead. Many (if not most) don’t even handle the actual voice data that travels to and from the end-users in an attempt to minimize overhead costs and maximize business solvency with very low pricing to the consumer. Adding a 64.9% theorized interstate revenue adjustment to small VoIP companies based on sheer guesswork and estimation, knowing full well that smaller companies can scarcely afford to pay salaries, much less commission FCC-approved studies for traffic flow, is going to put smaller VoIP companies in a position of being unable to compete.

With the global nature of the Internet, many small VoIP companies are taking advantage of the increased userbase that the world offers, allowing origination and termination in any number of countries. This regulation, however, would purport to offer that most voice traffic is US-centric and interstate and therefore subject to paying into the USF — a situation which is highly unlikely, and tantamount to taxation of non-US citizens, as any additional costs will have to be passed on to the userbase as a whole.

2. What significant alternatives should the FCC consider?

The FCC seems wholly unwilling to consider any alternatives until they can work out any issues they have with actual distribution of the collected funds — something that seems beyond the purvey of the FCC. If they have, until now, had problems with distribution, it is unlikely that in future they will be able to better regulate themselves. Such a decision will be external.

A solution the FCC could consider would be using less guesswork and more data in their arrival of a safe harbour number for VoIP providers. Alternatively, and more appropriately, the FCC could simply charge a fee based on every DID in the US local area to pay into the USF — bypassing the need to create traffic studies of any kind or to assume that all voice traffic is US-based.

3. Can wireless providers determine actual interstate and international end-user revenues.

This seems reasonable, considering they know which tower a phone is connected to (needed for authorisation on the network), as well as to which tower the call connects. There should be an easy one to one mapping for the purpose of creating absolute data.

4. Should originating and terminating cell sites be used to determine the jurisdictional nature of a call.

If the attempt is truly to determine which calls are interstate in nature, then yes. If the attempt is simply to collect a phone tax, then it’s somewhat irrelevant.

5. Is 37.1% for a wireless carrier appropriate or should it be raised or eliminated?

I’m of the opinion that, because of the nature of what the USF is used for, the rate should be eliminated.

Think, for a moment, back to what the USF is for. The USF is given to rural areas and areas with poor communications infrastructure in order to pay for communications infrastructure. VoIP companies and Wireless providers usually don’t have penetration into the rural areas, leaving the only ones out there to be the Wireline providers. What’s happening is that the wireline providers are paying the least into the USF, but getting a portion of that money back simply because they refuse to enhance the quality of the network infrastructure in rural areas. Meanwhile, VoIP companies and Wireless providers are, in essence, subsidising the wireline infrastructure in rural areas, thereby subsidising their competitors. The USF is clearly NOT suited for a non-homogeneous wireline telecommunications infrastructure. When you add VoIP and Wireless into the mix, it becomes severely anticompetitive.

6. How should the FCC set safe harbour percentages to better reflect market conditions?

Relying on a more accurate study than the TMS Telecoms study would be a good start. The FCC claims that, because of long-distance service bundling, wireless and VoIP companies are more likely to be doing interstate traffic than wireline providers. It is quite rare, however, to find a wireline provider who doesn’t also offer bundling. If it’s good for the goose, it should be good for the gander. The only way to fairly apply a USF tax is to make it a flat number across the board.

7. How can requirements on interconnected VoIP providers be improved?

Requiring a VoIP provider to pay into the USF for data that clearly never touches the PSTN until it hits an endpoint is a mess. That’s much akin to asking people to pay postage for email. While some would like it (most notably the USPS), it destroys the concept of a data stream for which one already pays. End users pay for bandwidth. VoIP companies pay for bandwidth. These payments may be flat rate or per megabyte charges, but the payments are still made, and somewhere along the line, those payments go into the pockets of companies that provide infrastructure for either telecom or data networks.

Where does it end? Is everyone along the line guilty of somehow subverting the USF system? Should everyone pay into the USF from the provider on down to the end user? How does one say that the provider of what amounts to being an automated phonebook for VoIP users is the one who should be paying into the USF to subsidise the wireline companies in rural areas? It seems to be a rather less based on how the technology works and more simply a line drawn in the sand at random.

8. Should the FCC change or eliminate the safe harbour for VoIP providers?

I believe there are other methods by which a more appropriate USF tax could be collected, but I still disagree that VoIP providers should be responsible for paying into a fund that goes to their direct competitors (wireline providers). If, however, the USF must be collected, and the competitors of VoIP must be governmentally subsidised, then the USF should be a flat safe harbour for all involved. No one method of communication is more or less guilty of providing interstate communication, and to assume otherwise is to penalise providers based on outdated and unproven assumptions.

9. Can VoIP providers identify actual interstate revenue?

Until VoIP providers are able to collect accurate and untampered-with IP data from all customers (difficult in the age of IP spoofing and redirection proxies), and until that data can be accurately mapped to an area (while the Geo-IP database is a handy tool, it is in no way perfect or 100% accurate), then no. There’s simply no way to accurately identify actual interstate revenue, and to create numbers — 23%, 31.7%, 64.9% — is to simply create an arbitrary system based on guesswork.

Infineon Technologies AG, your friendly-neighborhood German technology company, has decided to patent a method by which VoIP technology could be secretly degraded over a network. This would allow phone companies to intentionally make VoIP calls for their competitors not work very well, but wouldn’t cut off the ability completely, thereby making it seem as though the competitor simply has poor quality or service.

The patents, here and here cover traditional broadband and also wireless networks.

The technology works by inserting artificial data into a call that looks and behaves like VoIP data, but only serves to make a connection congested with false data. The data could then be removed to cover one’s tracks so as to avoid detection by the VoIP provider.

It’s nice to know someone is hard at work coming up ways to ensure that new, useful technology fails in spectacularly anticompetitive ways.