I first received the article ‘Why EDGE versus 3G matters less than you think’ on July 1 from a good friend, the day after I had signed up for a 1 year plan on the Safaricom Internet plan that uses a HSDPA USB modem on their 3G network. He had signed up for the Celtel EDGE Internet plan two days prior.
The article troubled me, but I was rusty on the technology that enables Internet to be transmitted via mobile phones or devices linked to mobile phone networks. So I decided to do some research and see if the author of the article was correct. Today I found a link to the same article on the post one of my favorite blogs, Bankelele, so I have decided to complete my research and offer my response to this odd article.
The author’s premise seems to be a defense to Apple’s decision to develop their initial iPhone to operate on EDGE networks rather than 3G networks. I will therefore assume and hope that this article is at least one year old since the iPhone 3G which will start selling in 20 countries around the world tomorrow, July 11 2008, has been developed for 3G networks.
Is there any need to respond to the rest of the article? Now that even Apple seems to have decided that 3G is better than EDGE.
I do not think so but for arguments sake let me continue. From my understanding in the first paragraph the author is based in the USA; ‘…here in the US …’. This is a very critical point to note as mobile phone networks in the USA are inferior to those in Europe, Asia and I dare say in many places in Africa. In all fairness to the USA most homes had very good land lines hence there was a smaller ready market for mobile phones and the USA is more sparsely populated than your typical European or Asian country so designing a good cellular network is more expensive and difficult to justify if you have a smaller ready market. 3G networks launched in Asia and Europe 2 years earlier than in the USA mainly because there was huge speculation on the services that could be offered on a faster network and data was the next opportunity for the mobile operators who were very eager to continue making high profits. I suspect that many in the USA would want to defend their cellular networks and show that they are not missing out on what the rest of the world has by arguing there isn’t much difference between EDGE and 3G. Would it be alright then to argue that there isn’t much difference between a GRPS network and an EDGE network?
Anyway to the ‘meat’ of the authors argument;
‘People confuse network bandwidth with latency’
That may be true but let us first define those terms and let me cite that my source for these and other definitions in these articles are from the amazing Wikipedia.
Bandwidth - a bit rate measured in bits/s, for example, network throughput (the average rate of successful message delivery over a communication channel).
Latency - the delay between the initiation of a network transmission by a sender and the receipt of that transmission by a receiver. In two-way communication, it may be measured as the time from the transmission of a request for a message, to the time when the message is successfully received.
Now in my understanding the two definitely affect the perceived fastness of an Internet connection and can be interrelated but allow me to state the following;
- Bandwidth, or throughput, is increased by technological advances e.g. multiplexing techniques such as Code Division Multiple Access (CDMA) and Frequency Division Duplex (FDD) to maximize the data transmitted. The 3G technology utilized by Safaricom is the High-Speed Downlink Packet Access (HSDPA). Please note that this does matter as the higher the bandwidth the faster it will take for data packets to arrive in a set time slot. However, please also note that the bandwidth often quoted for a technology, also referred to as the maximum theoretical throughput is the maximum possible but often much higher than what the normal throughput the user will have. Therefore a 3G network that may have a maximum theoretical throughput of 2mbps may only offer 236kbps but an EDGE user will not get their theoretical maximum of 236kbps but instead may receive 128kbps.
- All networks suffer latency, the primary contributors are; propagation (the time it takes for the packet to travel), the medium of transmission (a optical fiber link offers very little latency in comparison to a satellite link), the router and other processing and other computer and storage delays. It is debatable how much more latency is caused by the operation of 3G networks on a higher end of the frequency spectrum in comparison to an EDGE network.
‘Higher bandwidth radio networks are more error-prone’
I am not completely certain what the author means by a narrow-band radio, the Wikipedia definition is narrowband is usually considered to cover frequencies 300–3400 Hz. It is important to note that 3G networks are defined according to International Telecommunication Union (ITU) standards defined by IMT under the International Mobile Telecommunications programme, IMT-2000. To the chagrin of the ITU which had hoped that there would be one standard to end the madness of CDMA vs GSM, there are several technologies that are recognized as 3G. The common ones are CDMA2000 EVDO utilizing W-CDMA standard and the other is UMTS/HSPDA which is from the GSM ‘family’. The UMTS/HSPDA networks utilize the 850 MHz or the 2100 MHz frequency whilst the CDMA2000 operates at 450 MHz, 700 MHz, 800 MHz, 900 MHz, 1700 MHz, 1800 MHz, 1900 MHz, and 2100MHz. EDGE networks operate on GSM networks which operate on 850MHz, 900MHz, 1800MHz and 1900MHz frequencies.
‘Phone processors and software don’t necessarily keep up with fast data transmission’
There are two interesting arguments in this section by the author that have nothing to do with EDGE vs 3G. One is about the processor and software of the phone, the author suggests the Nokia feels slower than the iPhone. Really? The other is the odd statement about the ‘… 54 megabit WiFi – a network several times faster than that of the fastest 3G network’ I suspect that the 54mbps was the connection speed between the phone and the WiFi access point but the user will only be able to enjoy the Internet speed that the ISP is providing to the access point. If it is an optic fiber link of 2mbps that’s all the use will enjoy. But I suspect that that will seem slow because the user may be expecting 54mbps.
‘High bandwidth networks drain batteries’
Here I agree with the author that high bandwidth connections are power hungry and will drain batteries faster. But the fact that Apple has developed an iPhone for the 3G networks means that this hurdle can be overcome. The author also refers to the higher density of cell towers in Europe as opposed to the USA here.
3G versus EDGE does matter. The reason I believe it does is that when Hutchison Telecommunications were launching their 3G network dubbed Three in April 2003, I was a masters student in Australia. I remember listening to one of my lecturers, who taught us how to design telecommunications networks being very dismissive of 3G networks and any chance they had of providing reliable data services; Internet, TV etc. He made a compelling argument but perhaps because he kept referring to Zimbabwe, where he had worked for many years, as Rhodesia, I felt the man might not know everything and perhaps with continual improvements it may one day be possible to have Internet via 3G networks. I am posting this blog in Nairobi in my house courtesy of an Internet link on the Safaricom 3G network. It is one of the better deals at a rate of KES 2,000 for 700MB. That said I would be crazy for me to stay with a 3G network when it becomes possible to have a optic fiber cable to the home connection when Kenya becomes connected to the rest of the Internet using the undersea optic fiber cables in 2009 - 2010.