This week at the WiMax Forum Congress Asia in Singapore, SingTel’s CEO Allan Lew mentioned plans for a 40Mbps nationwide wireless boradband network in two years.
Considering I only have a 2Mpbs cable connection at home, HSDPA only at 3.6Mbps at max while on the go, and Wireless@SG is unreliable at best, the 40Mbps pervasive wireless network is most welcome.
In layman speak, for what some people call a 4G network, there are basically two very similar technologies that make sense here: 3G Long-term evolution (LTE) and WiMax. Both are very similar in design and pretty much serve the same purpose. So what will it be for Singapore?
I for one had my hopes up for WiMax since SingTel was already running mobile WiMax trials and Allan Lew did make the announcement above at WiMax Fourm, but it appears that WiMax may not be the ideal technology for a country like Singapore.
According to GigaOm:
Fred Wright, an SVP that handles 4G networks for Motorola, believes LTE will be the standard chosen by 80 percent of the carriers in the world — good news for vendors such as such as Alcatel-Lucent and Ericsson, who have opted to stick with LTE. Of course, as GSM is the dominant mobile standard today, such a prediction isn’t all that surprising.
In my opinion, mobile broadband services such as HSDPA have beat Wireless@SG’s sporadic metro WiFi as the connection of choice and going the 3G LTE route is an obvious upgrade for the operators instead of WiMax.
Other commentators also believe 3G LTE is better suited for mature markets (like Singapore) and WiMax for rural/emerging markets where no GSM infrastructure is in place. WiMax is a great technology because it can serve as back haul as well as access making it ideal for countries with larger landmass and huge rural populations.
But coming back to savvy Singapore, the rate of adoption is highly reliant on the availability of the end point devices. If Intel keeps to its promise and starts shipping all Centrino 2 Notebooks with WiMax this year, will users shy away from ugly LTE dongles/PC cards?
In a perfect world, I’d want both. Maybe different operators running different networks? But that’s unlikely.
Filed under: Singapore, Technology
[...] two years. Considering I only have a 2Mpbs cable connection at home, HSDPA only at 3.6Mbps at max whttp://eok.net/2008/04/12/singapores-4g-plans-lte-could-beat-wimax/Implementing a Highly Distributed, Real-time Data-critical SOA Application SYS-CON MediaSOA [...]
Dear Sir
I developed some algoritm to access 4G network withot any complexity. This algorithm was published at ieee digital library. If you allow me then i can establish those algorithm to you laboratory.
Yours sincerely
Ripon
As far as air-interface (which determines the capacity of the system) is concerned, LTE only has two advantages over WiMAX -
* Low frequency of operation
* SC-FDMA which improves uplink performance
These translates to coverage advantages, but not capacity.
Both LTE and WiMAX use OFDMA, MIMO, and TxAA; and follow almost the same set of numerology be it on the time or frequency domain.
Given the same amount of spectrum, both LTE and WiMAX are equivalent in terms of spectral efficiency (and capacity).
But what drives people to say that LTE has “better” capacity is because, in LTE, 20Mhz channel bandwidths are allowed. While 20Mhz is possible in WiMAX, only 10Mhz is certifiable.
So when it comes to capacity comparisons, 20Mhz LTE is compared with 10Mhz WiMAX, hence, LTE getting the upperhand.
But as far LTE spectrum is concerned, this involves frequency already utilized by GSM, and 3G systems which need to be refarmed for use with LTE.
Yet worldwide GSM, and 3G spectrum are normally allocated in smaller chunks, 15Mhz at most, and wont be enough to attain the capacity promise of LTE. At best, 5Mhz channels will be used in 3-sector configuration (similar to 3G), but this wont beat WiMAX operators who have ‘more’ spectrum, and who implement their systems in 10Mhz albeit at higher frequencies.
There are two ways to go around the spectrum dilemma in LTE -
(1) Deploy in previously unutilized frequencies – 700Mhz or 2.6Ghz.
700Mhz is only available in few countries. And 2.6Ghz is already allocated to WiMAX operators in a lot of places.
Even when these frequencies become available for LTE, it will need more spectrum than WiMAX to be competitive (given the fact that LTE will be on FDD initially).
(2) Operators share spectrum and RAN infrastructure. This is allowed in LTE via the S1-flex mechanism which allow eNBs (radio network) to be connected to multiple MMEs (core network).
But to have competiting operators share the same RAN just to attain the capacity benefits of LTE is politically difficult.
At the end of the day, LTE has the same capacity performance as WIMAX given the same spectrum, and unless regulators cough up the additional spectrum (which is near impossible) which drives LTE’s “better capacity”, you won’t see any tangible difference. And worldwide trials already show that.
But LTE remains to be the best choice for incumbents who already hold the spectrum LTE was built for.
As for operators who want the same performance but dont have these “incumbent frequencies”, they have WiMAX at 2.3Ghz, 2.5Ghz and 3.5Ghz.
3GPP folks should not be arrogant to say that only cellular incumbants are the only ones that should provide Wireless Broadband.
At the end of the day, the telecoms industry as a whole need to gear up for the next evolution, which is “Services Convergence”.
Customers are demanding voice, video, and broadband data from converged platforms – smart devices and new networks that need to support converged services (triple play + mobility).
To address this,
(1) Traditional ISPs, and Cable operators are upgrading their networks to bring REAL mobility to their existing broadband data offerings.
(2) Cellular incumbents are upgrading their networks to bring REAL broadband into their existing mobile services.
And these leads to 3.9G technologies that need to address network convergence at both the radio access and core network. (The industry has not technically agreed on what can be considered “4G”)
3GPP intends to tackle this convergence using LTE. HSPA and HSPA+ exists and may address the radio component of converged services, but certainly not with its outdated, complicated, and latency-laden core. Hence the need for LTE.
But LTE was really built for cellular incumbents – its frequency of operation, and core network depencies are designed to leverage on what incumbents already have.
LTE is intended to work over 2G and 3G spectrum, and nowehere in the world new cellular frequencies are being auction, let alone for new entrants.
What solutions do they offer then to ISPs, DSL, Cable operators and multimedia companies who have money but also want to provide converged services (Mobile Broadband) to their customers?
This is where WiMAX comes into the picture.
If LTE is the convergence solution for cellular incumbents, WiMAX is the equivalent for non-incumbents (over non-incumbent spectrum).
There is no such thing as WiMAX beating LTE, or LTE beating WiMAX, for that matter.
Vendors and operators who propagate that LTE is the only way to go does nothing but display sheer ignorance on where our telecom industry is really leading into.
LTE uplink based on SCFDMA is not an advantage over WiMAX, instead it is a huge liaibility.
UL MIMO works best with OFDMA on the uplink the receiver structures are very straight forward and you get best performance. This is what WiMAX has.
Instead by LTE choosing SCFDMA they are going to lose big time. They are struggling to meet Uplink performance numbers now….:)