Muni-Wi-Fi and Developments in the Device MarketWednesday, August 29, 2007
This week let's start with muni-Wi-Fi. The news is growing worse by the day. Chicago just announced it is cancelling its plans for a muni-Wi-Fi network because "it is too costly and too few residents will use it." Also cited was that it would take "massive public financing."
Meanwhile, one company that had been working on the Chicago system, EarthLink, announced it is cutting 900 jobs from a total workforce of about 2,000 and closing offices in various parts of the country. As part of the lay-offs, it has eliminated the position of president of municipal networks and laid off the person who filled this post.
Recently, there have been articles in many of the major news magazines about the perils of muni-Wi-Fi―most of them explain away the failures or stagnation of deployments as having to do with the cost of the systems. This is mostly likely true, and many cities that wanted to implement muni-Wi-Fi systems have refused to become the anchor tenant. Meanwhile, the cost of network deployments has gone up as the model for city-wide coverage has changed dramatically in only a few years.
These systems were designed as mesh networks and the original rule of thumb was that thirty access points were needed per square mile, and a backhaul connection was needed for every three access points. Fast forward to today and we find the recommended number of access points has more than doubled. I am told by those planning systems that they are now using sixty or more per square mile as the "norm," with even more in densely populated areas. Not only does this drive up the cost of deployment, it at least doubles the number of backhaul connections (either wired or wireless) that are needed.
All of this costs more money up front and increases long-term maintenance costs. Unfortunately, even with these additional access points being installed, in-building coverage is still almost non-existent. Meanwhile, the cost of cable and DSL continues to come down. Locally, a cable connection of 1.5 Mbps down and 256 up runs $14.95 a month for six months. It's not a great connection, but from what we have seen in our own muni-Wi-Fi tests, this is about the average speed available on a muni-Wi-Fi system with little or no loading.
Since these articles began appearing in the business and news magazines this month, I have received a lot of email commenting on how I have been predicting this muni-Wi-Fi meltdown for more than two years. Looking back at my Commentaries, they are right, I have been, but it is not because I have a crystal ball. It is because I run the numbers, understand that unlicensed spectrum cannot be usurped by a city and understand that as Wi-Fi grows in popularity within homes and businesses, the amount of interference to a muni-Wi-Fi system will continue to increase.
Look at T-Mobile's new Hotspot @Home system. It has tightly integrated Wi-Fi access points for use inside homes and businesses with its wide-area network. This gives T-Mobile the ability to offer coverage inside buildings, add value to its network and transfer some of the traffic on the wide-area network to access points that cover small areas inside homes and offices. T-Mobile won't be the only company to roll out this type of dual coverage.
Another thing we are seeing is that more and more free hotspots are being added all the time. The local laundromat now has free Wi-Fi for use while you wash and dry your clothes, and of the four coffee shops and restaurants in our little sub-community, there is a Starbucks with pay-for Wi-Fi and three others with free Wi-Fi. All of these, whether for-pay or free, are using Wi-Fi as it was intended to be used―as a local, in-building wireless connection to the Internet and on to your email and office (with the proper VPN or security software).
The muni-Wi-Fi bubble has burst. Those who tried to mine the gold of muni-Wi-Fi systems and cities that thought they could provide reliable broadband services to the masses are discovering the dream of broadband everywhere is not going to be realized with muni-Wi-Fi systems. Would the same thing work if Wi-Fi was replaced with WiMAX, and 2.4 unlicensed spectrum with 2.5 or 3.5 licensed spectrum? I believe it would. But the cost of this type of deployment is too great and voice is still what pays the bills and keeps the lights on for network operators. Data still accounts for less than 20% of their revenue.
The Sprint/Clearwire approach of providing wireless broadband coverage to a host of Internet appliances might be getting a boost from the demise of muni-Wi-Fi, but again, it will depend on coverage in buildings and not just on the streets of a city. It will also depend on who is right about what the wireless Internet should look like. The majority opinion at this point seems to be that it will merely be the wired Internet moved to a smaller, wireless device. The minority opinion, which I share, is that we need to use our smart wireless networks and smart wireless devices in a more efficient manner. Combining the smarts in the network with the smarts in the device would result in a great user experience and really fast access to data and entertainment options.
This brings me to my next topic: smartphones vs. small computers that fit in our pockets. What will we be carrying? There are many very robust smartphones already on the market, and the iPhone certainly fits into this category. However, I find one of the most interesting new devices to be the Nokia N81 and N95. If you read the press releases from Nokia and Symbian, you might notice what I did, which is that they are not calling these devices smartphones, they are calling them multimedia computers. The descriptions for both of these devices use the term "computer" over and over again. The N95 is a really high-end device with built-in A-GPS, a 5-megapixel camera, both GSM/UMTS/HSDPA and Wi-Fi built in and what looks like a great 2.8-inch screen with Quarter-VGA resolution. The only thing missing for the U.S market is the ability to receive the new AT&T TV channels via Qualcomm's MediaFLO network. Even so, the N95 is a full-blown computer with 8 GB of storage on board. While it will probably sell for more than $700 when it comes to the United States, this is in line with pricing I have seen for WiMAX ultra-portable devices.
When you look at the specs for the new Qualcomm Snapdragon processor, you will find that it is a 1-GHz microprocessor paired with 128-bit SIMD (Single Instruction Multiple Data) capability and a 600-MHz digital signal processor. It can be used on the CDMA2000 1xEV-DO and UMTS/HSDPA/HSUPA networks and it includes support for both Bluetooth and Wi-Fi as well as four different mobile TV broadcast standards (MediaFLO, ISBD-T, TDMB, DVB-H). And this is only the first of such chips to see the light of day.
With devices such as the Nokia N95 multimedia computer and those that will be built using the Snapdragon processor, the computer industry, which has long believed it will drive wireless adoption of the Internet by shrinking the notebook and expanding the capabilities of the PDA, will discover that the wireless industry also has designs on this space. At the moment, it looks to me as though the wireless side is ahead.
Wi-Fi will return to what it is best at providing, which is local-area wireless connectivity, wide-area systems will continue to incorporate Wi-Fi or roll out femtocells to provide in-building coverage and wireless device vendors will ramp up wireless computers. But there is still plenty of room for computer companies to play in this space.
I use two guidelines to categorize devices. The first is whether it will be used for information access in the field or for access and work creation. If it is for access only, the device can be smaller, while if it is for work creation, as in report writing, spreadsheets, etc., a larger device with a bigger keyboard best suits the task. The second is what I refer to as a conscious/unconscious carry. The unconscious carry device is one you always have with you, such as your phone, BlackBerry or Treo, while a conscious carry device is one you have to consider whether you want to take with you. For me, that means if I am going on a short trip I might leave my notebook at home and take only my phone and BlackBerry, but if I am traveling for more than a day, my notebook goes with me.
Networks are getting smarter and devices are getting smarter. How we take advantage of this combination will impact how many of us spend more money on wireless information and entertainment. It will be exciting to watch as mobile devices continue coming to us from both computer and wireless companies. We will have a wide variety of device choices, but the smartest device in the world is merely a brick unless it is attached to a smart network to empower it.
Andrew M. Seybold