Category: Uncategorized

In the standards, LTE has been specified for speeds up to around 350 km/h. The faster a user is moving the more tricky it gets to deal with fading, Doppler effects and handovers between cells. But that speed value from the standards is just a theoretical value, to be proven in practice.

It seems Huawei has just done that when they recently demonstrated their network's capabilities to journalists during a super high speed magnetic levitation train ride from Shanghai to Shanghai airport. At speeds of 430 km/h the connection was still maintained at high throughput as reported here by CNet news. Interesting pictures of the event are available here.

Via: Lightreading

Already since the very early days of 3G, there have been 5 air interface states in the specs: Cell-DCH, Cell-FACH, Cell-PCH, URA-PCH and Idle. In all networks I have so far encountered, only Cell-DCH, Cell-FACH and Idle have been used. I haven't seen the other two states yet, which can reduce the time it takes to return from a dormant to a fully active state, so I could only speculate on their effectiveness. Until today.

Here is a typical round trip delay (ping) behavior when returning from Idle to Cell-DCH:

PING 85.214.10.20 1000 (1028) bytes of data.
1008 bytes from 85.214.10.20: icmp_seq=1 ttl=47 time=2571 ms
1008 bytes from 85.214.10.20: icmp_seq=2 ttl=47 time=226 ms
1008 bytes from 85.214.10.20: icmp_seq=3 ttl=47 time=236 ms

Due to the state switching the answer to the first ping, the request takes over 2.5 seconds. Especially when web browsing this delay is quite noticeable.

And this is the round trip delay profile when returning from Cell- or URA-PCH to Cell-DCH:

PING 85.214.10.20 1000 (1028) bytes of data.
1008 bytes from 85.214.10.20: icmp_seq=1 ttl=47 time=987 ms
1008 bytes from 85.214.10.20: icmp_seq=2 ttl=47 time=236 ms
1008 bytes from 85.214.10.20: icmp_seq=3 ttl=47 time=236 ms

Instead of 2.5 seconds, the first answer is already received in less than a second, which significantly improves the overall user experience. Further, I observed that the network is configured with a fully active Cell-DCH state for 20 seconds after which the Cell- or URA-PCH state is entered pretty quickly without staying too long in Cell-FACH state.

Once in Cell- or URA-PCH state the network first goes to Cell-FACH and remains there if only small IP packets are are sent. This was the case for example with a standard ping which is why for the example above, a 1000 byte ping message was used to trigger an immediate state change to Cell-DCH. This is quite typical for web browsing as when clicking on a link, the HTTP request package has that kind of size if the page is on the same web server and hence no DNS lookup is required, the TCP link is already established and the request including cookies are sent straight away.

I also did some tests with my somewhat 'older' N95 8GB and with an old firmware version. Here the first packet took around 1500 ms to arrive. Quite a bit slower than the times above reached with a current Class-7 HSPA 3G USB stick but still much faster than returning from Idle state.

For further background reading on the state changes and procedures there's always my book on the topic for a more detailed introduction and 3GPP specs TS 25.331 for all the details.

Being used to multi-megabit/s 3G speeds and a 25 MBit/s VDSL line at home I expected to have a real bad browsing and general Internet experience when I recently experimented with a 3G HSPA connection throttled down to 384 kbit/s in the downlink direction and 64 kbit/s in the uplink direction. I was quite surprised to experience the contrary.

Yes, the web pages are showing up with a bit more delay than I am used to but apart from that, the general experience was quite o.k. Youtube doesn't work though without buffering as the videos require a bit more bandwidth. Apart from that, just the 64 kbit/s in uplink hurt, uploading pictures took a while. But still, quite usable overall, much much better than I expected!

In case you have a good memory, the title of this post may sound familiar. Back in October last year, I wrote a post on a paper published by Ericsson in which they came to the conclusion that as networks evolve and demand for mobile Internet access rises, the cost of transporting one gigabyte of data over a wireless network can be in the range of 1 Euro. Now NSN has published a similar paper and come to the same conclusion. A very interesting read as they explain very well how they come to their conclusion. It's also in line with the calculations I made in my recent book and it's good to see that the assumptions I made at the time were also made by others to estimate capacity.

An interesting twist of the paper is that it breaks with this nice little graph you might be familiar with showing that as Internet use rises in wireless networks, there's a rift opening up between revenue and amount of CAPEX and OPEX if things left as they are today. The paper actually says that the reverse can  happen if network operators evolve their networks: The more the network is used, i.e. the more people pay a monthly usage fee for wireless Internet access, the cheaper it gets per subscriber to transport a certain amount of data for the network operator.

Finally, the German spectrum auction for everything between 800 MHz and 2.6 GHz assigned for cellular services has ended. While the media mostly reports the €4.4 billion as a disappointment for the German government, I personally think it's more than enough for a few pieces of thin air. And, as a side note, compared to the sums paid in the Netherlands and in the Nordic countries, it's a respectable sum. Now it's in the hands of the four incumbent mobile network operators to make the best use of their new spectrum in the coming years. For the details on who has acquired what have a look here.

Dean Bubley over at Disruptive Wireless has recently published an interesting post in which he wonders whether he has or will have a problem with Wi-Fi and femto offload in the future.

He describes how he is walking from his home to the tube station and his mobile device in the pocket picking up several known Wi-Fi hotspots from BT and connecting to them for a couple of seconds just to loose them again. This wreaks havoc on applications that are actively exchanging information over the network even when the phone is not actively used as the connection is dropped every time. An interesting point he makes there and I don't have a single and simple solution for it but some things come to my mind on how to deal with this:

On the Wi-Fi side:

• The connection manager could wait for a couple of seconds checking the signal strength and only connecting when it stays relatively constant, i.e. the user is not moving and hence, the Wi-Fi hotspot is worth using.

• Distinguish between public Wi-Fi hotspots where such rules should be applied and encrypted home / office hotspots for which it could apply different rules.

• A good Wi-Fi offload solution would be not to get a new IP address in every new hotspot and to break the 3G connectivity. Ideally, an IP tunnel is used so no matter whether connected over 3G or Wi-Fi the device would always use the same IP address. That would also deal with the security issues of unencrypted public Wi-Fi hotspots.

On the Femto side:

• I think there would be less of a problem here because GSM, UMTS and LTE offer myriads of possibilities to ensure a mobile only uses a femto if it is not moving.
• There is no need to assign a new IP address when hopping between femtos so connectivity would not be broken when jumping into or out of a femto.
• The connection can be handed over from and to a femto so continued connectivity with little or no outage can be ensured.
• Access to a femto can be restricted. This book goes into the details of this.

So, yes, small cells are a challenge for mobility but there are options to deal with this.

In the days where notebooks were too big to carry around with, too expensive to buy a smaller one just for the purpose and too heavy for the bag I tried to get as much done as  possible with the mobile phone. Writing a blog entry on the phone, a section for the next book or a lengthy e-mail with a foldable keyboard  was practical but never quite enjoyable.

It's not that the user experience could not have been better but the software never developed much beyond it's original incarnation. Over the years I never got a spell checker, no 'app' today to make blogging a bit easier from Typepad and viewing PDF documents on my mobile remains a pain to this day. While it remained the only choice it was better than just keeping my thoughts for later.

The netbook I bought last year, however, pretty much changed the game for me. With a small battery and weighing less than a kilogram, my netbook comfortably fits in my bag, resumes from suspend in just a few seconds and the screen size and processing power is sufficient to comfortably do all the described tasks above plus more. Even opening it up for just 5 minutes to do a quick task is viable and 3G plus Wi-Fi connect me instantly.

That doesn't mean of course that I no longer carry a mobile phone, I just use it less for content creation but more for consumption oriented tasks these days such as web browsing, feed reading and for getting instant e-mail notification. Ah yes, and for phone calls of course. That makes me think that now that the keyboard has become less important I might try a touch based phone again.

A quick post today to make those of you who are interested in the finer details of LTE aware of an interesting post by Zahid Ghadialy over at the 3G and 4G Wireless blog on LTE conformance testing. Towards the end of the post, Zahid links to a couple of simulator test cases developed by 3GPP RAN5 and the GCF for which actual logs are attached. In other words, the theory you can learn from books is taken into practice there and you can see how the signaling messages look like between the mobile and the network and what their contents are. Great stuff, thanks Zahid!

23 auction days, 186 rounds and there is still no end in sight in the German Spectrum auction for 800, 1800, 2100 and 2600 GHz frequencies.

While for some time, there was hectic bidding for spectrum in the 800 MHz digital dividend band, things seem to be settled there for the moment as there hasn't been any movement there since day 16. If things remain there as they are right now, all four network operators will own spectrum for around €420 million per 2×5 MHz slot. Vodafone and T-Mobile each hold slots each while Telefonica/O2 and E-Plus
(KPN) each hold one slot. Out of the €3.2 billion currently in the pot, €2.5 billion are for the 800 MHz band. In other words, all the rest can be had for cheap.

Since bidding in the 800 MHz band has halted, things have been trickling along. Since day 16, the total sum has 'only' increased by €0.4 billion. The auction will only resume next Monday due to a public holiday in Germany on Thursday. So how much longer will it still drag out? I hope not too long anymore…

Ever since I read this book on how mobile networks have spread in Africa and discovered Eric Hersman's blog on technology and life in Africa in the process, I've been following his stories of what's going on there. In case you are planning a trip to Kenya and wonder how to best stay connected to the Internet, here's a recent post of Eric in which he describes how he keeps connected. Should be applicable to other countries in Africa as well. Makes for an interesting read and reminds me of the early days of GSM and GPRS in other parts of the world!