WirelessMoves

Here I am, back in Italy and again using my 3 UK SIM for Internet access, since there are no data roaming charges between 3 networks in different countries. Very interesting and also a bit depressing to see the performance throughout the day. Sunday morninig 8 am seems to be a pretty quiet time and I easily got 1.5 MBit/s in downlink. Evenings seem to be high time, with data rates dropping to less than 300 kbit/s and long page loading times due to lots of lost packets. I am pretty sure it’s not a cell overload since the Wind UMTS network at the same locations easily gives me 1 MBit/s and no packet loss in direct comparison. So the bottleneck is either the link back to the home network or the GGSN in the UK .

While it’s good to see the networks being used and affordable data roaming in place, I’d appreciate sufficient capacity in the core network.

Since I travel a lot, I often stay in hotels. One thing in hotel rooms I could live without is the TV set, as I never have the desire nor the time to watch anything anyway. In recent years, however, I've noticed that good old cathode ray tube TV's are giving way to TFT TV's. Last week I took a closer look and noticed that these TV's usually also have a VGA or DVI input. Excellent, now I can finally put them into good use and connect them to my notebook as a second screen. All that is needed is a VGA or DVI cable, which I will take with me from now on. The picture on the left shows how my typical hotel setup looks like. Screen 1, screen 2 and a 3.5G HSPA modem for Internet connectivity. Almost as good as at home 🙂

In a comment to a recent blog entry somebody left a link to an interesting S60 utility that records power consumption and other interesting technical information while Noka NSeries and other S60 based devices execute programs. Running in the background, one can perform actions and see the impact on power consumption. The picture on the left for example shows power consumption during different states during mobile web browsing.

The left side of the graph (click to enlarge) shows power consumption while a web page is loaded, i.e. while the device is in Cell-DCH / HSPA state. Power consumption is at almost 2 watts during this time. Once no more data is transferred, the mobile is set into Cell-FACH state by the network which requires much less power. However, the 0.8 watts is still significant. After about 30 seconds of inactivity the network releases the physical bearer and only maintains a logical connection. In this state, power requirements drop to about 0.2 watts which is mostly used for driving the display and the background light. When the device is locked and the background light is switched off, the graph almost drops to the bottom, i.e. to less than 0.1 watts.

An excellent tool to gain a better understanding of power requirements of different actions and processes!

I am on the road quite often and, as most of you have figured out in the meantime, a heavy user of 3G networks for Internet access. While I generally like the experience some network outages like this two and a half day nationwide full Internet access blackout in the Vodafone Germany network recently sends shivers down my spine. After all, we are not talking about a third class operator but one that claims to be a technology leader in the sector. As I use Vodafone Internet access a lot I was glad I was only impacted for half a day, having been in a DSL save haven for the rest of the time. If I had been on the road, however, this would have been a major disaster for me.

I wonder if the company that delivered the equipment that paralyzed the Vodafone network for two and a half days has to pay for the caused damage!? If it’s a small company then such a prolonged outage with millions of euros in lost revenue can easily put them out of business. And that doesn’t even consider the image loss for all parties involved and the financial loss of companies relying on Vodafone to provide Internet access. The name of the culprit was not released to the press but those working in the industry know very well what happened. Hard times for certain marketing people on the horizon…

Vodafone is certainly not alone facing such issues as I can observe occasional connection issues with other network operators as well. These, however, are usually short in nature and range from a couple of minutes to an hour or so. Bad enough.

To me, this shows several things:

• There is not a lot of redundancy built into the network.
• Disaster recovery and upgrade procedures are not very well thought trough as otherwise such prolonged outages would not happen.
• Short outages might be caused by software bugs and resetting devices.
• I think we might have reached a point where capacity of core network nodes have reached a level that the failure of one device triggers nationwide outages.

So maybe operators should start thinking in earnest about reversing the trend a bit and consider decentralization again to reduce the effect of fatal equipment failures. And maybe price should not be the only criteria to be considered in the future. Higher reliability and credible disaster recovery mechanisms which do not only work on paper might be worth something as well. An opportunity for network vendors to distinguish themselves?

Recently I spoke to a sales engineer of Kathrein, a prominent German antenna maker that develops and produces all sorts of antenna equipment from TV antennas to sophisticated cellular network antennas. I can still remember how simple antennas were 10 years ago when GSM was first deployed: In many places simple dipol antennas were used and sometimes funny looking trident antennas (that’s how I call them anyway, I am sure they have a more scientific name…). In the meantime we have mostly moved to bipolar antennas that offer a main and diversity output to the base station. On top of that, lots of other things have been developed which are either now being deployed or waiting for that 4G bandwidth push that requires sophisticated antenna features:

Dual band antennas, e.g. 900 + 2100 MHz in one standard casing for sites with GSM and UMTS base stations: Such antennas give themselves away with four connectors at the bottom.

Wideband antennas, e.g. 2.1 – 2.5 GHz to support UMTS and LTE with a single antenna: I am sure those will be in high demand once LTE is deployed in the 2.5 GHz range.

Cable reduction: To reduce the number of expensive coax copper cables between the base station and the antennas, combiners have been developed to combine the signals of several base stations, send them through a single cable and then separate them again before they go into the different antennas.

Remote Electrical Tilt (RET): The size of a cell mostly depends on the angle of the antenna at the rooftop. The more it is tilted towards the ground, the smaller the coverage area. When a new base station is installed to increase the availble capacity in an area, it is necessary to change the tilt of neighboring antennas to reduce interference. Also as capacity in the network increases, it is also sometimes necessary to change the tilt of antennas to improve the overall coverage and bandwidth distribution. Manually changing the tilt of an antenna for these scenarios is expensive and sometimes simply not possible. This is where RET comes in. Instead of physically changing the angle of the antenna, RET changes it by increasing the lengths of the different antenna elements inside the antenna casing. This way, the overall antenna can be electrically tilted by around 10 degrees. Practically, changing the antenna lengths is done with an electrical motor that drives a spindle inside the antenna casing which can change the antenna lenghts of the different modules. The electrical motor is an add on module at the bottom of the antenna.

Antenna auto adjust: It can also be imagined that RET is used in the future to automatically ajust antenna angles based for example on the time of day. This could help to increase coverage in certain areas at certain times of the day by decreasing the cell size. Things could be pushed even further by linking the RET mechanism to the load of the cell and increase the tilt when the cell gets busy to offload some of the borader traffic to neighboring cells.

MIMO and Beamforming: And then there is Multiple Input, Multiple Output and beamforming for further bandwidth increases that requires several antennas at the rooftop. In practice, they are again included in a single casing to look like a single antenna to the outside. The first 2×2 MIMO systems will use little crossed antenna elements that send each MIMO channels with a vertial and a horizontal polarization.

All highly interesting and shows how important antennas have become for increasing bandwidth in the future. Thanks to Kathrein for the interesting information!

I’ve been in Rome many times before and the main tourist attractions have lost a bit of appeal to me. So I decided to discover some of the more hidden gems on my current trip. Discovery, that is the difficult thing for me, since it’s nice to read something in a tourist guide but quite another to know how far that attraction is from the hotel or the place I am at the time. So I decided to try a Nokia Maps City Guide plugin. For Rome, three were available from different sources. As each is usable for 10 minutes before one has to pay for it, I downloaded all three and finally decided for WCities Guide.

When the plugin is started, one can choose from a number of categories. I was interested in sightseeing so I chose that. Next, a list is presented with the sights closest to the current position on top. They are also shown in normal map mode so one can easily see where the different sights are in relation to the current position and get some first info by clicking on the icons. Very nice! That makes it very simple to decide where to go next and how to get there. There is only little textual information provided for each location but with the help of the Internet, OperaMini and Wikipedia it is quite easy to get full background information and pictures on almost anything.

Here’s an example: I decided to visit San Clemente, a basilika close to the Colosseo which is built on top of a fourth century church which is built on a house dating back to the Roman empire. So after finding the place with the City Guide and Nokia Maps, I started OperaMini and used Google to find the relevant Wikipedia entry on the San Clemente Basilica in Rome. Fantastic, lots of background information in there and since it is all linked you can venture out and discover the life of San Clemens, San Cyril and other people connected to the location. Better than any general tourist guide book! The rest of the day went accordingly.

From the above it is quite obvious that I really liked the experience and I had a great day walking through Rome and discovering things this way I haven’t seen before. Naturally, there are some things that could still be improved. Here are a few:

The WCities Guide doesn’t precisely pinpoint the location of many sights they feature. That made it sometimes a bit difficult to find the place, as in a city, even an inaccuracy of a hundred meters or so puts you into an entirely different place.

I’d really like a clickable link in the City Guide description of a location to Wikipedia or another encyclopedia for further information. Ideally, this would open up the web browser. I could also live with a copy/paste feature for the URL, that is much better in combination with OperaMini anyway.

Price: 8 euros for a guide to a single city is a bit steep. After all, no book has to be printed, there are no shipping and storage cost, so I don’t quite see why I should pay almost as much as for a book.

Altogether, a great solution for a different sightseeing experience and I will surely use it again. Also nice to see where things could go with this in the future. Lots of potential!

Today, we don’t think a great deal when we pick up a mobile phone to call somebody while in the car, or a train or moving otherwise. The network, if properly designed, ensures that the connection is handed over from cell to cell as we move. With most phones, this works rather well, probably because the protocol stacks have been refined for this over many years and it is tested intensly. After all, it is the main purpose of the device.

With the rising use of mobile Internet connectivity, one would assume the same would be true for devices that have an IP connection established while being carried around. However, over the past few years, I have observed now that this is not really the case and I have to admit that I am a bit frustrated by hanging and rebooting mobiles while being connected to the net while moving. At first I thought this was due to the immaturity of the technology. However, UMTS networks are around for several years now and the situation has not really improved. Here are some examples:

My Sony-Ericsson V800: This is the first 3G mobile I bought several years ago and I never managed to get hold of a firmware that would not reboot when the mobile was used in a train as a data modem. O.k., it was one of the first usable 3G phones so I give it the benefit of doubt.

The Motorola V3xx: I use this device exclusively as a 3.5G data modem and in general it works quiet nicely. It’s quite stable even in a train or other moving vehicle but does reset at least once an hour, too when moving.

The Nokia N95-8GB: This one is a notorious ‘rebooter’. As soon as I launch OperaMini and move, even when just on foot, it reboots…

So this rebooting while connected to the packet network during cell reselections is not a one company, one device issue, it seems to be a universal phenomenon. It shows that comlexitiy has risen beyond a point where such things can be tested well enough before devices are shipped. Or is it just the companies that don’t (yet) care?

But there is a little hope: Today, I tested the Nokia N82 and OperaMini while in the car and it didn’t reboot like the N95. That’s good since this device just went into hands which are even less forgiving when things don’t work.

This February at the Mobile World Congress, Intel invited me to experience their WiMAX network that they had set up in the city during the event. Of course I had lots of technical questions and one of them was on the backhaul equipment they used to connect the base stations to the network. I was told that they were using 50 Mbit/s microwave links with equipment from Dragonwave. Impressive I thought at the time and was thus happy to see that Dragonwave attended a recent WiMAX event in Munich. An excellent opportunity to get more details. Turns out that the 50 MBit/s, hyperfast compared to the 2 MBit/s E1 links used today in most 3G networks, is just the entry level speed. According to the data sheets, the system is capable of doing about 800 MBit/s with dual polarization and 256QAM modulation over a 56MHz channel and 1.6 GBit/s when using 2 channels. The range of the system at this speed is several kilometers. That's much faster than what you need even for LTE and WiMAX in the short and medium term per base station. However, base stations are usually daisy chained in a ring configuration so if you chain five base stations together that each require 100 MBit/s (e.g. three sectors), that's already 500 MBit/s.

The longest distance they have bridged with their kit so far at 50 MBit/s hiltop to hiltop was 75km. Another impressing number.

Equally impressing is the power output required. I always thought microwave equipment operates at very high power. Looks like I was wrong, the output power of the system is a mere half watt. Of course the gain of the antenna is very high but still…

As the system transports Ethernet/IP frames natively, most LTE and WiMAX base station can use the system right away. Current 3.5G base stations, however, are still using TDM links. For these, a pseudo wire box can be used to tunnel such connections over Ethernet.

I also learnt a lot about Microwave frequency license costs. Looks like the cost differs widely. In the U.S., I was told, a basic license for a channel (can't remember the exact bandwidth anymore) is around 1800 dollars for ten years. Compare that to France where the cost for the same license is around a thousand Euros  for a single year. Channel sizes that can be licensed start around 6 MHz and go up to 28 MHz. Again depending on the country, the microwave spectrum is anywhere between 11and 38 GHz.

Like all wireless systems, microwave links don't like things such as rain and snowfall. E1 based microwave links therefore need a security margin in order not to fail under such conditions. In the world of Ethernet, however, this margin can be translated into higher speeds when conditions are fine and the automatic link adaptation during  rain and snowfall by automatically lowering the data rate. By prioritizing VoIP packets, it can be ensured that voice calls are not impacted by this while background priority traffic such as web browsing gets slower during such times. I know, easier said than done in practice, IMS definitely needed for this. But that's another topic.

When asked on the price point of the system I was quoted a mid four digit number for an end to end link, depending of course on the quantity purchased. Compared to the price of a single 2 MBit/s E1 link, which easily reaches several hundred Euros per month, that's very competitive. Comforting to know that backhaul prices scale well with rising consumer demand for wireless access.

Thanks again to Dragonwave for the interesting interview, I learnt a lot!