Networking: Focus on Cellular Telephony (Mobile Phones)
Sunday 28 June 2020
Dr R. Haines
Back in Networking Lesson 4 (Transmission Media), we looked at different wireless transmission media: Wi-Fi, Bluetooth, satellite and cellular. This lesson / web-page is extension work, a “deep dive” into cellular telephony to enrich your STEM education, give you some context and a look at the wider world of this topic, and maybe even pique your interest enough to take this topic further through studying Engineering at college and university.
To be really clear: all you need to know about cellular telephony to pass your GCSE in Computer Science is more than covered in Lesson 4 (Transmission Media) of the Networking topic (or this BBC Bitesize page): the syllabus does not explicitly ask for you to know anything about cellular (sadly!) – although I would argue that you need to know how cellular is different in order to understand Wi-Fi better, and to understand what your ‘phone is doing.
When your phone or tablet is not on Wi-Fi and is “using your data” while on 3G / 4G / LTE / 5G, it is connecting to the cellular mobile telephone network.
The mobile telephone network is very different to Wi-Fi or Ethernet, even though it gives you the same services and applications (e.g. web browsing).
Wi-Fi and Ethernet came from the computing industry, the companies that build computers and servers and printers and LANs, so they’ve been designed from the beginning to connect computers together – computers that are often plugged into the mains too.
Cellular has come from the telephone industry – think of table-top telephones with handsets, wires, buttons or dials (landlines); switchboards and big national organisations such as BT and AT&T.
So, even though both end up doing a similar thing as far as you (as an end-user) are concerned, in the background they work and are designed in a very different way. Cellular uses different technologies, at different frequencies, and, rather than being run by your company’s or school’s IT department – or yourself if you’re at home – it is all planned, built, managed and maintained by the Network Operator (e.g. Vodafone, EE,…), which is one reason why you have to pay for it (compared to just getting onto “free” Wi-Fi).
Another reason is the radio channels used. Wi-Fi and Bluetooth all jostle for space in free, noisy, unlicensed band (e.g. 2.4GHz ISM, 5GHz UNII). Cellular, however, operates in entirely different bands (900MHz, 1800MHz, …) that have been auctioned off by all the governments around the world, paid for (costing millions if not billions in each country) by the network operators (Vodafone, EE, TIM, Sprint, etc.) and are now privately owned, for the sole use of each network.
So, there’s no fighting in the band with other systems, there’s much less interference, and the operators are in control: they tell your device what frequency to go on, how to connect, what data rate it’s going to get, whether it even can connect, and they can fine tune it all to work like clockwork.
In this lesson/blog, we’re going to have a look at the history of cellular (“how did we get here?”), what the main parts are in a cellular network, a quick look at one example of how it works (just to show you how complicated it is in the background!), and then finish by looking at the main differences between cellular and LAN/WLAN.
First of all, we talk about “being on 3G”, EE are advertising their 5G networks … what does that “G” stand for? Surprisingly, as Michael Caine would say, not a lot of people know that.
G is for… Generations
The very earliest “mobile” telephones were available just after World War II (using some of the radio technologies and equipment developed for the military) – for example, the “Mobile Telephone System” in St Louis, Missouri, USA, launched in 1946.
Each “telephone” weighed over 80lbs (36kg) – that’s about the same as an average Year 7 pupil 🙂 Unsurprisingly, you didn’t carry them around with you, they were permanently fitted in cars, like your car stereo is today.
To make a call, you had to speak to the operator (a human!) and tell them what number you wanted to ring. They’d then place the call and connect you to them.
Initially, the entire city of St Louis was covered by just three (3) radio channels, so if you were the fourth person who wanted to make or receive a call at any particular time, tough! Also, it was entirely possible that a car that was closer to the the main transmitter/receiver (transceiver) in the middle of the city would steal the channel, because that car’s radio signal would be much stronger (because it is closer) than a car further away.
The system could not handle many users – but that was not a huge problem as it was so expensive that it was only the richest people in society that could afford it anyway!
At time went on, more channels were added and systems spread across the USA (and similar systems grew in other countries, e.g. the Post Office Radiophone Service launched in the UK in 1959). This led to another problem – neighbouring cities couldn’t use the same frequencies, as they would interfere with each other. Also as time went on, the equipment got smaller until it would fit… in a suitcase (wow!!).
These systems (and improved versions, such as the “Improved Mobile Telephone Service”, which that phone-in-a-suitcase above is an example of — note that customers can now dial the number they’re calling themselves, progress!), believe it or not, were still in use as recently as the 1990s.
The big problem with the 0G systems was capacity – the systems simply could not support the number of people that wanted to use them. Customers would have to wait ages (e.g. half an hour) for a free channel so that they could make a call.
The solution was cells. It’s the same word as in biology, and if you think of what onion cells look like through a microscope you are on the right lines. The idea is to divide each area into smaller areas (classic computer science decomposition!). Each area can then have calls running at the same time as the neighbouring cells, providing each cell is using a different set of frequencies. The smaller the cells, the greater the capacity of the system (and cost, as the company has to build more and more base stations, a transceiver for each cell, connecting the radio telephones to the telephone network).
You can even re-use frequencies – as long as the cells using the same frequencies are far enough apart to not interfere with each other (e.g., in the picture above, look at the red cells – they never touch, and there other cells between them).
Originally, if you moved out of one cell into another during a phone call, you would get cut-off, until the idea of handover (UK) or handoff (US) was invented which meant the system could quickly transfer your phone call to the next cell, and your mobile telephone would jump from cell to cell like Spider-Man swinging across Manhattan from building to building.
Mobile phones were now becoming (almost) recognisable – on 3 April 1973, Martin Cooper of Motorola made the first ever mobile telephone call from a handheld phone, and, in an a marvellous example of throwing some shade, made that phone call to Dr. Joel S. Engel of rival company Bell Labs. This photo is of him re-enacting it over 30 years later:
As you can see, the phone was not exactly pocket-sized: it weighed over a kilogram, and after half an hour’s use, you had to charge it for ten hours.
These phones got smaller and smaller, and different standards sprang up all over the world – AMPS in the USA, TACS in the UK, NTC in Scandinavia, C-450 in West Germany… Japan had three completely separate systems at this point! These were the classic “yuppie” phones of the 1980s (watch the movie Wall Street when you’re a bit older) – large enough to be used as a defensive weapon, and working double duty as a dumbbell when in use. By the early 1990s they had shrunk down to a comparable size to today’s mobile phones – below is the Nokia 232, which weighed just ~200g. Alicia Silverstone uses the American (AMPS) version in the 1995 film Clueless (which also stars a very young Paul Rudd, aka Ant Man)… and in the UK, my wife’s Nokia 232 was the first mobile owned in my family.
(As a brief post-script, incredibly, there was still a 1G system working, an NMT network in Russia, in 2018! Let that be a lesson to you: your software or designs might be in use for a lot longer than you expect!).
One of the problems with 1G was, as I mentioned in the previous section, that there was a different standard in almost every country, which is a pain if you want to move from country to country. While much of the story so far has taken place in the USA, we now switch to Europe, where the European Union was forming, so the idea of having just one standard for the whole of Europe was really attractive. The European telecommunications organisations (CEPT) and later the European Telecommunications Standard Institute (based in the south of France – but not in one of the nice bits, it’s hidden in a forest in the middle of nowhere… you keep expecting Ewoks to jump out at any moment) formed the Groupe Spécial Mobile (GSM) to come up with this standard. Later, GSM was retconned to stand for “Global System for Mobile Communications“.
Also, there were now new digital technologies that could be used to replace the analogue technologies in 1G, which would mean we could squeeze more users into the same amount of radio bandwidth (up to eight times more with GSM over TACS), and do lots of other fancy tricks like encrypting the calls to stop people listening in, power saving to make the battery life last longer, and even be able to send written messages (you know, messages you actually type in to your phone on the keyboard, using text) and connect to this new-fangled Internet thing that people are beginning to use at home – although those last two ideas probably won’t take off 😉
To borrow from the Bee Gees, this is where I came in, my second job after graduation was with NEC, designing, writing, testing and debugging software for GSM mobile telephones, such as these:
2G phones introduced text messages (SMS), international roaming (where you can take your phone abroad and it’ll still works), SIM cards, and the ability to connect to the Internet.
The world’s first ever text message was sent on Vodafone by a British engineer called Neil Papworth in 1992, and as companies got their phones designed and working, each could claim to have sent the first SMS on one their company’s phones. A couple of years after Mr Papworth’s text, your humble author, having got NEC’s second-generation 2G hardware up and running, debugged the device driver software, installed the protocol stack, and sent the first text message any NEC phone had ever sent. #ClaimToFame.
2G phones represented a lot of extremes – the nigh-indestructible Nokia 3310 has spawned a whole series of memes celebrating its legendary battery life – measured in days, not hours – (e.g. this or this) and how tough it was (e.g. this). Companies also competed to make the smallest phones they could: NEC produced the world’s first <100cc phone, the DB2000 – a fact celebrated by an example of that phone being in a display case (middle shelf, on the left) in the National Museum of Computing:
(As the DB2000 has my software in it, that officially makes me a museum piece…)
2G got better and better at doing data – the original GSM spec could do 9.6kbps, the GPRS (General Packet Radio Service) additions, launched in 2000, could do up to 40kbps, and the EDGE (Enhanced Data rates for GSM Evolution) extension in 2003 took 2G (often called 2.5G because of these additions) up to 384kbps. So next time you’re complaining because your mobile has only got GPRS or EDGE and is streaming your TikToks really slowly, remember that the technologies were invented before you were born, cut them some slack!
Much of the world – even beyond Europe – was using GSM, with a couple of exceptions (e.g. the USA, who were mostly using a completely incompatible CDMA-based system, with a few GSM networks here and there), so there was a push to do what GSM did in producing a Europe-wide standard, but this time produce a world-wide standard, a Universal Mobile Telephone System (UMTS), which we sort of managed (with a few regional variations).
In the time since GSM had been launched (early 1990s), the Internet had gone from a home hobby curiosity and academic communications system to being something that could make money and that the general public wanted to use regularly – the first online shops were appearing (e.g. Amazon in 1994), and even early social networking (e.g. Friends Reunited in 2000). So, 3G put a lot more focus on data from the outset (with GSM, it had been a belated add-on), aiming for “mobile broadband“: offering 144kbps – 384kbps when on the move, and up to 2Mbps if you were stationary (not moving).
3G also saw the arrival of video calling, a short-lived cut-down Internet called WAP, proper Internet browsing and email access, the ability to play music (MP3s) and phones that look more like phones do today (more screen and fewer buttons or buttons that at least flip out of the way).
By the late 2000’s, it was clear that even 3G was going to struggle with all the streaming video and audio that customers wanted to do (I’m looking at you, YouTube and Spotify!).
The big change in 4G is one that you’ll never notice as a user. Up to GPRS (2.5G), everything was circuit switched – if you were making a call or connecting to a server, you had a logical channel, a circuit, dedicated to you for the whole of that call. The core network is acting like an old-fashioned switchboard operator, physically connecting one caller to another.
GPRS brought in packet switching (like the way IP packets are routed over the Internet) for the air-interface (mobile to base-station link). 4G went the whole hog and removed all the circuit switching in the core network too: now everything was like the Internet: packets being routed and switched. If you hear the term “VoIP” in terms of mobile phones, this is Voice over Internet Protocol – i.e. making that packet switching feel as if you have a circuit switched permanent telephone connection to the person you are talking to.
The main “4G” (there’s some philosophical debate as to whether it is true 4G) is LTE – Long Term Evolution (so named because this big jump to packet switching was seen as setting us up for the future). At the radio-end, LTE ushered in data rates of 10Mbps – 100Mbps (300Mbps theoretically) and lower latency (delay – an important factor if you are playing games on your mobile: if you tell your Fortnite avatar to jump, you need it to jump NOW, not in 2 seconds’ time!).
In 5G, we’re looking at even greater data rates (up to 2Gbps), new frequencies (e.g. 60GHz millimetre wave and even, to the horror of the Wi-Fi community, sneaking into the unlicensed 2.4GHz and 5GHz bands) and new technologies (e.g. MIMO, see the deep-dive Wi-Fi blog, specifically the 802.11n part) but also new ways to configure and optimise the networks to support new applications, such as the Internet-of-Things (also known as machine-to-machine, M2M) and dealing better with big crowded events with tens of thousands of users (ever tried making a call at a football match or concert? Pre-lockdown, of course…)
And no, for the record, there is no evidence that 5G does anything harmful to you – it is non-ionising radiation at the long (radio) wavelength end of the spectrum, not the nasty high-energy high-frequency stuff at the gamma / X-ray end. The frequencies being used today were being used in the 1980s with 1G and 2G systems, which used much higher power levels than today’s devices – if there was some massive health implication (e.g. brain cancer), we’d be seeing it by now (like you could see the widespread cases of lung cancer 20-30 years after smokers started smoking in the middle of the 20th century). And no, a radio wave can’t carry a biological virus – that’s scientifically impossible, go and ask a science teacher (e.g. me!). Again, just think about it logically: there were Covid-19 outbreaks in countries without 5G, whilst New Zealand, which does have 5G, had the lowest number of cases and deaths (about two dozen in total) imaginable. This article by Ofcom also clears up some of the myths. Science 101: if someone makes a huge claim (e.g. “5G causes Covid-19!”) then they need to back that up with evidence. As it says on the wall of my science classroom (from 2018 to the time of writing at least), nullius in verba – ‘take nobody’s word for it‘: always demand to see the evidence, perform the experiments and make up your own mind. It’s the motto of the Royal Society in London.
Extension task: draw a timeline of the evolution of the cellular generations – include the approximate dates, the possible data rates (when applicable) and any other information you want to add. Can you also add any wider historical context – what was happening in the world at different points along that timeline? When were you born? How about your parents or carers?
Quick Task: What do you think are the different parts of a mobile network? What bits have you seen around town? If you’ve read the history section above, you should have had some hints. Once you’ve had a think about it, compare what you came up with with the information in the next section.
For Wi-Fi, you have a WLAN (Wireless Local Area Network) NIC (Network Interface Card) in your laptop or mobile phone, and that connects you to the WLAN AP (Access Point). The AP then either acts as a switch, connecting you to a LAN, or as a router, connecting you to a WAN (e.g. the Internet). And that’s about it, because after the AP, it’s someone else’s problem.
What you are getting for your money with cellular is an entire managed and optimised system. For the purposes of explaining this, I’ll stick to the 2G GSM structure because it’s a bit simpler, the blocks are the same (with slightly different names) even today, and it’s the one I know best!
Your phone! It’s actually two things: the “Mobile Equipment” (ME) phone hardware itself (e.g. Apple iPhone, Samsung Galaxy) and that little plug in SIM (Subscriber Identity Module) card. These two things together are called a Mobile Station (MS), just like the nodes in Wi-Fi are called stations. The reason there are two parts (ME + SIM=MS) dates back to GSM – in trying to make the system truly pan-European, the engineers in ETSI were worried that maybe some countries might have to use slightly different handsets or systems, but if we all used the same SIM cards, you could pop your card out from your phone when you got there and put it in a local rental phone, which would then connect as you. Also, if you rented a car (which could have a more powerful carphone built into it), you could pop your SIM in there and all your calls would follow you to that rented phone. That’s also why, even today, you usually have the option to save your contacts to your SIM – it’s so you can take them with you if you move your SIM to another ME. Also, fun fact, the original SIM cards were the size of credit cards, not the tiny finger-nail sized things we have now.
Next in the chain is the base station – actually comprised of the transceiver (base transceiver station, BTS) itself plus a base station controller (BSC) that could be controlling one or more BTS. This is the equivalent to the Wi-Fi AP – it’s a massive radio transceiver, talking to multiple mobile phones at the same time, carefully controlling power levels so that it can be heard by (and hear) distant MS at the same time as really close ones, helping MS handover to other cells, and handover into this one, and routing all the data into and out of the core network behind it. For some reason, for 3G, they started calling this a “node B”, which makes no sense to me and seems really inelegant linguistically. It might make sense if there was a “node A” somewhere (yes, I know the B stands for base station!).
Now we’re into what people working on the radio end of it all (e.g. me) refer to vaguely as the core network. This includes:
the Switching Centre – formerly circuit-based (following on from old analogue telephone systems), and now (mostly, depending on how far a Network Operator has got in rolling out 4G) packet-based, this is the junction where everything gets routed to the right place – your web browser requests are sent off to the Internet, your phone call to Domino’s is aimed at the telephone network, your text message to your mate on the same network is bounced toward another base station somewhere else in the network.
the Home Location Register (HLR) does 2 jobs. First of all, it remembers where you are (which cell you are in), so that, for example, in the case of that text message heading to your mate, it knows which base station to send it to. It’s also really useful when someone rings you: if the network didn’t know where you were, it would have to “page” you (effectively say “hey! you have a call!“) in every cell in the entire network, which would waste a lot of bandwidth. Secondly, it maps your phone number (07773 …) to your a unique ID number that your SIM has built into it (called an IMSI). When you get a new phone and they say that you don’t have to move your SIM across, there’s already one in it, what is happening is the HLR is being updated to match your phone number to the new SIM ID (IMSI).
Quick Task: Sketch the main components of a mobile network. If you want to practise your IT skills, do it using the drawing and Insert Symbol functions in PowerPoint.
How it Works: Example – Receiving a Phone Call
The diagram above is a Message Sequence Chart, which can be used in any communications protocol. Time advances going down the diagram (careful, that’s different to science where we usually put time on the x-axis. If you’re struggling, lean your head onto your right shoulder.). Each vertical line is a different “thing” on the network – on the left is a phone on the regular telephone network somewhere (e.g. it could be my mum ringing me up to find out where I am). Next we’ve got the mobile switching centre (MSC), then the HLR, BSC and BTS mentioned in the previous section, and finally, on the right is the mobile phone (MS) that the regular telephone is calling (e.g. my mobile).
Working down the diagram, the MSC finds out from the HLR which cell the MS is currently in. The MS is currently power-saving (screen off, most chips powered down, etc. – it was all that power-saving that led to the famously long 2G battery life). The network starts paging the MS (announcing that there is an incoming call), and the MS wakes up and says “OK, OK, I’m waking up” (the second message, “RACH”, on the right). The MS and network then go through a quite long sequence of messages (most of which I’ve not drawn!) to get set up for the call, before the BSC assigns a logical channel to the MS for this call. If anything goes wrong at any point in all of this (e.g. there are no spare channels in that cell), the call will drop and the caller (my mum, remember) will just get diverted to voice-mail. If all has gone well then finally the MS will start ringing – until now, there’s been no indication to the user (e.g. me) that anything is happening. Assuming I answer it and don’t dump (“UDUB” – User Defined User Busy) my mum to voice-mail (as if I would!), the voice/audio paths will then get set up and away we go: “Hello there!“
Quick Task: Copy the message sequence chart above, make sure you note when the mobile actually starts ringing. Stretch/challenge: Draw a flow-chart for one or more things in the message sequence chart (e.g. the mobile phone). Remember to use the right symbols, including the terminators (start/stop sausage), decision diamonds, etc.
Final Task… extended writing activity, answer this Big Question:
“Compare and contrast Wi-Fi and cellular as wireless connection services.“ [6 marks]
Plan it: when faced with a 6-mark question and all that blank space, break it down. Say what you see (e.g. in the question), state what you know about the subject, describe what is happening (e.g. in the question) and then (FTW) explain how what you know makes what you have described happen. Write a PARAGRAPH like in English or history (PEEL etc.) Keywords: wireless, licenced/unlicenced, landline, broadband, base station, mobile station; Expressive Language: replacement, infrastructure, connect, guarantee; Connectives: via, therefore, which, because
Example answer… (no peeking)
Wi-Fi and cellular both provide wireless connectivity, for example, to the Internet. Wi-Fi comes from the IT industries, and provides a cable-replacement service for a LAN at the link / MAC layer. It operates in unlicensed radio frequency bands such as 2.4GHz and 5GHz. It usually connects to a wired LAN or, through a router, to the Internet. Cellular comes from the telephony world, initially providing the ability to make and take phone calls whilst away from a landline. It has now evolved to offer mobile broadband over a much advanced version of that infrastructure. Original 2G worked in the 900MHz and, later, the 1800MHz bands, but other bands have since been incorporated. The core network (specifically the switching centre) routes data to the Internet (and voice calls are routed to the landline telephone networks, other mobile networks or within the current network as appropriate). Wi-Fi is a best effort system where the system will try to get your data across the link, but there could be congestion or interference that stops it, because the radio bands are free for other systems and applications to use. Cellular offers more guarantees because the radio bands it uses are licensed and controlled by the network operators. Cellular also covers a much greater area (each GSM cell can be 70km in diameter – a WiFi AP will struggle to cover 50m). However, apart from the initial purchase of equipment and the monthly broadband bill (which can be shared among many users), Wi-Fi is largely free; cellular requires a monthly subscription with a contractual commitment, or regular top-ups through a Pay-As-You-Go arrangement, both of which are much more expensive in the longer term.
Or something like that – make sure you have stated some facts about each system (e.g. operating frequencies), and covered where they’re similar (e.g. both offer wireless data) and different (e.g. frequencies or architectures), and, for serious kerching! marks, given some pros and cons (e.g. Wi-Fi is free/cheaper… but cellular is much better quality).
Remember, you’re looking to give the examiner six reasons to give you a mark (the question was worth 6 marks), so go for the classic state/describe/explain or, if it’s a compare-and-contrast like this, make sure you throw in some facts about BOTH things, and give some pros/cons about each.