The Evolution of Mobile Data

2G, 3G, 4G- What is the difference?

It may be a bit hard to believe these days, but back in the day cell phones were actually called "cell phones", and not mobile phones, smartphones, superphones or feature phones They used to sit in our pockets and were meant to make calls - that was it. Yep, none of that messaging, browsing, Facebook, Twitter, Flash 10.1 rubbish. They couldn't upload 5 megapixel pics to Flickr, and they definitely couldn't create a WiFi hotspot.

Fortunately, those dark times are behind us; however, as providers around the globe are "cooking up" all kinds of up-and-coming next-generation high-speed wireless networks, it's getting a bit hard to keep track of it all. What does "4G" mean, for example? Sure, it's the next step from 3G, but does that automatically make it better? And why did all 4 major national carriers in America start calling their networks 4G all of a sudden? Maybe it's all the same? To find answers to those questions we'll have to take a walk through the historical dynasty of wireless networks - past, present, and future. Don't worry - it'll be a most enjoyable journey.
Note: the "G" means "generation", so, when folks talk about "4G", they actually refer to a wireless network based on 4th-generation technology. Alright, let's walk down the memory lane - right back to when the very first gen wireless network emerged on the market.

1G

Our trip starts in the early 80's, that's when a lot of innovative network technologies were introduced: AMPS in the States and a combo of TACS and a NMT in Europe. What those acronyms mean is not really important - no quiz later, I promise. What you should know is that these brand-new standards were better than their predecessors in that they had a big enough spectrum for pretty heavy customer use, they were completely automated on the providers end - there was no need for any interventions by human operators - and they used electronic components that were small enough and could fit into smaller packages (take a look at Motorola DynaTAC, for example). And even though there were several generations of mobile services prior to AMPS, TACS and NMT back in the 50's, the first generation (1G) is mainly associated with the introduction of those technologies, because they were the first who managed to make cell phones practical and useful for the general public. They were dependable, robust, and, over time, conquered pretty much the whole industrialized world and covered the needs of every advanced nation.
The fun part is, no one thought about data services when 1G first emerged; it was an analog tech that was developed for voice calls only. True, there were modems that could send data over 1G networks - certain handsets even had them built in - but, because analog cellular connections were far more noise-sensitive than regular landlines, data transfer was super slow. And it wouldn't really make a difference, if they were actually fast enough; the AMPS network's rates were ridiculous in the 80's and, therefore, cell phones were considered a luxury and were used mainly only by Wall Street execs, not the middle-class people. Moreover, they didn't have fancy smartphones that could download big chunks of data. Oh, and there was no YouTube either. The starts were just far from being in a perfect sync, I guess.

2G

Let's jump to 1993 - that's when Telecom (they're called Telstra now) developed the digital network. This new technology overcame a lot of issues with the AMPS network we just talked about, network congestion (overload, rather) and security being the two major ones. The 2G technology introduced many of the services we use today - SMS, multimedia messaging, access to the Internet, etc; plus, that's when SIM cards first came around.
This advanced new digital network was called GSM, and its technological foundation is TDMA (it's like FMDA). The GSM (Global System for Mobile Communication) operates on the radio frequency band of 900MHz spectrum; soon it made its way to the 1800MHz band as well.
So, how exactly is it better than AMPS? The answer lies in TDMA (Time Division Multiple Access). FDMA divides the 900MHZ (890MHz to 915MHz, to be exact) band into 124 individual channels - each 200KHz wide. Next, every single channel gets split into 8 0.577us bursts, which tremendously increases the maximum number of simultaneously connected users. We don't hear countless voices stacked up thanks to the miracles of digital compression, but we're not going to talk about that now.
Besides having more users per one cell tower, the 2G tech provides a lot of other useful features:

- 1: Digital encryption (64bit A5/1 stream cipher) 
- 2: Packet data (used for MMS/I-net connection) 
- 3: Text Messages - SMS
- 4: Caller ID along with other network features alike.

So, any problems? Of course! GSM has a very small coverage range (not a problem with AMPS). The downside to the TDMA technology (the core of the 2G network) is that if a cell phone can't respond within the pre-determined timeslot (0.577 us bursts), the cell tower will "drop" it and move on to the next call. Another vulnerability of  GSM is that the packet data transmission rates are very slow (like really-really slow), and if you're on a Vodafone, 3, Virgin, or Optus network then you're probably experienced this first-hand when you went outside the "coverage zone".
In order to get rid of these two issues we're going to take a look at the next 2 networks - CDMA and EDGE.

CDMA

The Code Division Multiple Access branch of the 2G tech first saw the light of day in 1999. Telstra introduced it as the next big thing for users who could receive good signals on AMPS but were outside of the GSM-defined range. The wide range is due to replacement of the "time" multiplexing with code-based multiplexing. In addition, the frequency band was lowered - to 800MHz - and that also helped in extending the range through reduced path loss and attenuation.
Imagine a room crowded with people talking - with TDMA each person waits for his/her turn to talk (that's time division); CDMA, on the other hand, makes it possible for multiple folks to talk simultaneously, but it's like every person speaks a different language (in a unique code that is). Obviously, that's not literally how it all works, so, if you want to learn more, check out the various resources at the bottom of this article.

EDGE

GSM developed a GPRS-based packed data network back in 2001; the maximum speed was between 60 to 80 kbps (downlink), with a download speed of 10kB/s - that's a bit faster than dial-up. Later on, they introduced EDGE - Enhanced Data Rate for GSM Evolution. It was a bolt-on protocol (that means no new tech was required), and it managed to boost the data rate of the 2G network up to 237kbps (29kB/s).
3G - A True Revolution
Three Mobile, together with Telstra, breathed life into the 3G standard in 2005, with the introduction of the 2100MHz network; initially serving major metropolitan areas, they soon expanded their coverage to 50 percent of the entire Australian population. The combination of the brand-new 2100MHz with the good-old 900MHz network, leased out to the major providers - Optus, Vodafone and Virgin - laid the foundation of all the non-Telstra broadband mobile services; today 3G serves around 94 percent of the Australian folks.
The 3G standard makes use of a new technology - UMTS (Universal Mobile Telecommunications System) - that's its key network syste,. This network successfully puts together some developments from the 2G network with brand-new technology and protocols to provide a rather impressive data rate.
At the heart of UMTS is the WCDMA air interface that has some similarities to the older CDMA in terms of technology, in that multiple users are able to transmit data on the same frequency through a code-based multiplexing. Wideband CDMA (WCDMA) uses this same concept and broadens the frequency band to 5MHz. This new tech also involves great algorithmic and mathematical improvements in signal transmission, which makes it possible to have more efficient transmissions at a significantly lower wattage - 250mW, as compared to 2W with 2G networks.
The 3G network also utilizes a significantly better (secure) encryption algorithm during over the air data transmission. 3G operates on a 128-bit A5/3 stream cipher that, unlike the A5/1 used in GSM (which can be relatively easily cracked near real-time with a ciphertext-only attack) has no weak points at all.
How Is 3G Faster Than EDGE?
UMTS uses the HSPA (High Speed Packet Access) protocol that's a merge of the HSDPA - downlink - and HSUPA - uplink - protocols. The Telstra HSDPA network works with category 10 gadgets (with a max speed of 14.4Mbps); however, the majority of devices can only transmit category 7-8 (7.2Mbps down), and the HSUPA network supports category 6 - that's 5.76Mbps up. These protocols are pretty sophisticated and they have a perfected transport layer thanks to a complex layout of physical layer channels (HS-SCCH, HS-DPCCH and HS-PDSCH). The technological side of the HSPA is too "techy" for us to get into right now, but if you want a basic description, watch the video below.
So, the only major downside of the 3G network is - obviously - coverage. As mentioned before the 2100MHz network is available to half of the Australian population, and, in combination with a 900MHz UMTS network it's available to almost everybody - 94%. As predicted, the 2100MHz network suffers a lot more attenuation and FSPL and is often referred to as a "short range" network; that's why the lower 900MHz network is required to cover most of the regional and rural areas.

4G - LTE-Advanced

Telstra's 4G network was introduced in October 2011, and, at that time, was only available in major cities, certain regional areas, and airports. It provides lower latency, very impressive speed, as well as reduced network congestion.
The 4G network is constructed upon LTE-Advanced - 3GPP Long Term Evolution. LTE is a pack of upgrades to the UMTS technology and will operate on Telstra's current 1800MHz frequency band. This brand-new network offers great advantages, with a download speed of up to 100Mbps and upload speed of 50Mbps, along with a reduced latency (100ms vs. the original 300ms), and significantly lower congestion.
4G bandwidth (the width of available frequencies that we can send/receive on, that is) is essential for providing high speeds for a large number of users. This is how it works: each individual user gets a small bunch of frequencies that they (and they only) can transmit on - that's how your connection doesn't get confused with another user's. There's a downside: during hours of heavy usage - meaning when lots of folks start connecting via the tower - the network will reduce everyone's frequency range, which will result in a slower download/upload speed.