Evolutionary Path



3G 4G and 5G (yes and they are talking 6G too)





Mobile phone technology has been on a steady path of evolution since the release of the first “1G” analogue phones. But just as it was with landline services, the transmission of data over the mobile network has been an important add on. Initially data use was a small component, but today, with the explosion of broadband services, mobile data is a key driver in the ongoing evolution  of the technology. 3G networks current dominate the global market, but here in Australia most Telcos have announced the time-line for the closure of their 3G services and are migrating to 4G. But while this is happening, there is a flurry of government and media interest in 5G. As this happens it is important that, as users of a small sub-set of this technology (the market for M2M or machine to machine communications) that you have a clear picture of what each generation of equipment can do for you
Australia’s last 2G systems were shut down soon after the rollout of the 3G network in 2012.





The launch of the third generation of mobile technologies came at a time when there was also rapid growth in the market for M2M systems. This started with the conversion of fire alarms and building security systems from traditional landlines to mobile phone based systems. Mobiles offered a lower operating cost and much lower installation costs: no longer did cables have to  be run from the building’s PABX cabinet to the alarm systems. Cellular or mobile modems could just be dropped in to the alarm cabinet and connected to power.
While previous generations of modems had been dial on demand, 3G heralded the widespread use of permanent connections. This parallels the shift in home Internet connections from dial up modems to broadband connections. With a dial up system, the modem would dial the Internet Service Provider (ISP) each time they wanted to access the Internet. They would then be charged according to the amount of time they spent on line. A broadband modem by comparison would remain permanently connected and users were charged for the amount of data they sent, not how much time they spent on line.
This change was a primary factor in the growth of M2M systems: while sending data over a dial up system may cost hundreds of dollars a year, the use of a permanent connection dropped this to a few dollars.
3G data systems also offered considerably higher speeds than their 2G predecessors. But the primary purpose of 3G systems was still the making of voice calls. Data was still consider d as an add on service. The future however ws being seen by broadband users who were able to switch from a traditional analogue landline phone to a service provided over their Internet connection: something known as VOIP of Voice Over Internet Protocol. Early services performed very badly, with frequent dropouts and speech delays. But, as the speed of the underlying network improved, VOIP phones offered a way of having a landline phone at a much lower cost than with a standard landline service. The technology behind VOIP has now matured to the point that as Australia has rolled out its National Broadband Network (NBN) customers can no longer have a landline phone and must convert to a VOIP phone.





While 1G, 2G and 3G were primarily voice services, with an added layer for data, the 4G system turns this on its head, with data being the prime system and telephone calls being made using VOIP technologies. Given the above mentioned transition in home and business Internet connections, this change seems a logical progression.
The 4G mobile system offers data rates which are considerably higher than those possible with 2G and 3G systems. These come about due to a number of factors, one of which is improved modulation techniques (which allow more data to be crammed in to a given transmission) another is the availability of additional bandwidth, which enables each channel to be wider and to hence carry more data.
However the largest contributor to the higher data rates is the pairing of frequencies:on a typical 4G implemention, devices can transmit on two different bands at once, hence doubling the throughput. In Australia, Telstra operates their 4G pairing on 700 MHz and 1800 Mhz. There is however one very important consideration here: the lower frequency 700 MHz services are capable of traveling much longer distances than the 1800 MHz service, which also suffers from poor ability to penetrate through obstacles such as buildings. So when deploying their dual frequency networks, providers rely on making the cells closer  together: instead of towers being 20 or 30cm apart, they will be 3 to 5km apart. This high cell density offers another advantage: if the total capacity of a tower is fixed, then by reducing the size of its coverage it will take less users, which in turn means that each user can send more data.
Those of you who work in country areas will have witnessed first hand the major consequence: if you visit a town, you will get high speed Internet via use of the 700 MHz and 18000 MHz bands. But as soon as you move out of the town boundaries, the 1800 MHz service drops off and you are back to 700 MHz alone. This means that rural users are automatically denied one of the biggest benefits of the 4G system.



Cat M1 and NB1


The mobile data systems have evolved to the point where they can now form the basis for many user’s high speed Internet access. But what about the segment which does not require high speed or high data volumes? This is the Machine to Machine or M2M market, where energy efficiency is usually more of a concern than speed
For years, the M2M market has been well served by the use of standard 3G and 4G data modems. The fact that the modems are operating well under capacity does not matter: they are cheap and reliable . Further, M2M data plans have been available at reasonable cost.
The evolution of long range, license free radio systems for M2M applications started to challenge the cosy position of the mobile carriers: suddenly LoRa WAN, SigFox, Weightless and the other new players could offer systems which did not rely on a “Telco”, were low cost and far more energy efficient. The threat of losing sales in a market segment which, as a result of the emergence of the “Internet of Things” motivated the mobile data industry to come up with an alternative.
Just as has been the tendency with most technological developments, there are two platforms competing for this new “Narrow Band Internet of Things” market: NB1 and Cat M1
There are advantages and disadvantages of each. But in Australia, one factor is driving the market to Cat M1: it does not require any new hardware and can be added to Telstra’s 4G towers with a software upgrade. Competitors offering NB1 must by comparison, install new hardware to support it, which means a longer, more complex and more expensive implementation path.
Modems which support Cat M1 and NB1  and approved for use in Australia  are now readily available. Manufacturers have chosen two different paths : they will offer a 4G modem which supports standard 4G as well as Cat M1/NB1 or offer a universal Cat M1 / NB1 modem. The first approach offers users the ability to switch easily between services according to the application. But it requires manufacturers to make multiple modems so that they can fit in with a country’s band plan. Offering a Cat M1 / NB1 only modem means the modem can be sold anywhere in the world. But the problem at present is that these modems do not offer 4G fall-back - so users still need to choose a different modem for applications which require a higher data rate.
With all the discussion of Cat M1 and NB1 it is important to realise that they have a number of deliberate limitations. Taking Cat M1 as an example, as that is the system most likely to be encountered in Australia:
Cat M1 does not use the timing advance function used on standard  4G VOIP and data channels. It thus does not suffer from the enforced 35km range limit. This is evident in the Cat M1 coverage maps which show some very long ranges: basically if you can capture signal you can communicate. The promise of Voice over M1 handsets will for the first time offer users the type of range they had 20 years ago with the first CDMA systems
Cat M1 is very low speed: this means that transmissions that took a fraction of a second on a 4G modem can take 20 seconds on Cat M1. This means you need to allow a longer time for each transmission. It also means that over the air firmware updates and sending of configuration files will take much longer
Cat M1 is intended for infrequent, short transmissions. This is how the system gets its promised reduction in capacity: some of the normal requirements for regular communications on the paging channel are relaxed, allowing units to go in to deep sleep mode between transmissions, helping to reduce power consumption
Cat M1 data rates will support  devices such as cameras, but are not suitable for video of data streaming.
Cat M1 devices are expected to drop off after 300 seconds, so systems which offer features such as keeping devices on line between polls will no longer work.
Of all the reasons to look at Cat M1, probably the most significant is its range. Normal 2G/3G/4G services carry an enforced range limit of about 35km. That is because the towers use “time division multiplexing” to increase capacity. Because radio waves take a finite time to travel, signals from devices close to the tower arrive before those from devices further away. So the towers allocate 64 individual time slots and each time you move another 500m from the tower, you get handed over to the next slot. Ideally when you get to the last one, you can be handed over to another tower. But if there is no tower, you drop out, no matter how strong your signal. The CatM1 services do not use this “Timing  advance” function and hence are free from the artificial range limitation. So if you can get your antenna mounted high, you can achieve very long range. Telstra’s coverage maps show many sites extending 60 to 100km, which means you will get Cat M1 coverage in many locations which previously had no service. Once they enable emergency dialing on Cat M1, you will be able to make emergency calls on the service when at your remote sites as well.





5G is being touted as the future and a lot of 4G users are worried about the prospect of 4G systems being closed down to make way for 5G (as it has with CDMA, 2G and 3G). If 4G marked the switch from systems based on voice to those designed for data, 5G marks the transition between low speed and high speed (in relative terms).
But physics plays an important qualifying role here. To increase data rates, designers must use more bandwidth. But since bandwidth is limited in the traditional mobile phone bands, spectrum managers are allocating new bands to mobile services on much higher frequencies. At the higher frequencies, each channel can occupy more space and hence transmit more data. Techniques such as band pairing can also be used to double throughput.
But gaining in one area always causes a compromise in another: the higher you go in frequency, the shorter the transmission range for a given power. Longer distances also mean higher error rates, so to reduce error rates, shorter range is preferred. Plus, if you have users demanding more and more data (to watch video on demand) if you only have a limited capacity, by reducing the system range, you can offer each user more data.
The net effect of all this is a move towards “Small Cells.”  With 5G, cells will be place 500m, 1km or 2km apart. The cell towers can be smaller and cheaper to deploy, but to cover a given area, that means a massive increased number of cells.  So no matter what the hype says, 5G is going the be limited to large urban centres for quite a while. Even the much vaunted push for self driving cars falls victim to this: as soon as 5G services cease, so too will the functionality need to keep the cars mobile.
What this means is that (1) 5G will not offer anything for rural customers for a long g time and (2) 5G offers absolutely nothing for the M2M market of which we are all a part. But since 5G systems will be rolled out in tandem with 4G systems, we will be able to rely on 4G - in its standard LTE and Cat M1 (LTE M1) guise -  for many years to come





Although the first practical 5G systems are only now being deployed, manufacturers are starting to develop (and promote) 6G. With 6G they are promising end user speeds of 1 gigabit per second ( 1 GB/s). The realists among you will straight away recognise the trap here: if increased data rates need increasingly high frequencies and high frequencies mean shorter range, then 6G is going to be even shorter range than 5G. Early 5G prototype base stations were the size of a mini-van. The first practical 6G base station will likely be just as big. But before they can be released they will need to be scaled down to briefcase size: otherwise Telco’s will have no way of installing them in the density they will be needed (i.,e. on a 500 to 1000m radius). So 6G too will have great promise for city dwellers but will be of no practical use for anyone shifting data in remote localities. And that after all is where we play.