LTE being a packet based technology; there have been ongoing discussions about Voice implementation over LTE. This is going to be crucial as operators would want just one network to be maintained to reduce their CAPEX and OPEX. There have been quite a few solutions being pondered upon in the 3GPP community.
One way is the CS fallback mechanism to address this. By CS fallback they mean that for voice LTE will handover the call to 2G/3G networks. This is done by MME registering the device with the 2G/3G networks and acting as SGSN communicating with MSC for voice signaling. This will ensure that the voice quality is maintained as of today, but this would also mean that if an existing packet data call is in progress, then it might get suspended or handed over to 2G/3G networks. I am not sure is this will be good idea since earlier generation networks will not have the bandwidth and the capacity to support LTE multi-media applications.
I read a paper about CSoHS (Circuit Switched over HSPA+) . This explains about how HSPA+ radio bearers will be used to carry voice traffic but core used will be CS core.
This can be done via software upgrade on UE as well as RNC. This would add de-jitter and time warping mechanism to ensure smooth voice quality. I hope this can be further extended to LTE, which is approach pursued by VoLGA which stands for Voice over LTE via Generic Access (Thanks to Steve for clarifying this!!). The difference between 2 approaches is avoiding the fallback to legacy networks and also using high capacity radio bearers. Also user not be required to suspend/handoff his data call.
Though these above approaches sound good for now, they would still require different networks to be maintained. Going further, IMS will have an important role to play as it can support end-to-end voice over IP services with call signaling and switching functionality required for voice. It also supports Voice Call Continuity (VCC) that can enable gradual deployment of VoIP.
Wednesday, June 10, 2009
Sunday, February 22, 2009
RevB Knocking on Carrier's Doors
Lately there has been a lot of buzz about LTE. Last week at Mobile World Congress 2009 in Barcelona, major wireless carriers like Verizon, Vodafone, AT&T, T-mobile announced plans to go with LTE as their next generation 4G technology. Verizon went even a step further to detail their vendors, commercial trials and deployment plan. So LTE looks all set to hit the trials by late 2009/early 2010.
Having said that I feel LTE is alteast 3 years away from commercial readiness. And here's why. If we take a deeper look into Verizon earnings call their CAPEX spending for 2008 was almost the same as their previous year. LTE deployment would require billions of dollars in CAPEX. Considering the current market economy, it is hard to imagine Verizon going with an aggressive rollout. So as per Verizon CTO Dick Lynch, we should see LTE rollout in 2011. But I believe it will be more of phased approach with all markets covered by 2013. We should see other carriers also follow the same approach.
LTE boasts of download speeds of about 34Mbps (I would assume it will be 10-15Mbps theoretically). LTE will be a platform for all data centric applications requiring higher speeds in the order of 8-10 Mbps. VoIP should also take off. But the phased approach in LTE deployment would require current 3G networks to support higher speeds. EVDO RevA can only support upto 3Mbps, though theoretically its must less than that (700Kbps-1Mbps).
Here is where EVDO RevB steps in for the 3GPP2 camp. For carriers like Verizon, it should only be a software upgrade not requiring much additional spending. 3 carrier RevB can support downlink speeds of upto 9Mbps and uplink speeds of upto 5Mbps. Qualcomm also announced availability of MSM8960 chipset supporting both LTE and EVDO RevB at MWC 2009. So if carriers decide to roll out RevB along with phased out LTE approach, they should see availability of dual mode handsets supporting LTE/RevB handovers with seemless mobility. This would enable applications designed for LTE getting traction soon resulting in rich user experience. Nobody wants their video conferencing session being closed just because they moved from LTE to EVDO network.
Now RevB makes even more sense to Sprint. The beleaguered carrier just lost 1.4M subs in Q4. WiMax even though being a good technology with time to market advantage, its facing the same destiny as EVDV. So even if WiMax is prevalent, I see it more of a backhaul solution than 802.16e. Moreover its now Clearwire's headache. So now Sprint is left with its EVDO RevA network. Their rival Verizon is moving to LTE. So they need a network that can compete with Verizon in terms of speed, supporing data hogging applications. And year on year there has been consistent increase in data usage. Cisco even predicts data usage going up with video services accounting for 64% of mobile traffic by 2013. Surely if the carriers are to support around 5GB demand for data per month they ought to have network that can meet the demand. With EVDO RevA I do not see that happening. Also the incremental cost of deploying RevB will be much less.
I think both Verizon and Sprint should go for RevB in order to get maximum out of their current 3G networks before we see 4G taking over. 3GPP camp will also do the same with Tmobile, AT&T upgrading their networks to HSPA + coupled with phased LTE rollout.
Having said that I feel LTE is alteast 3 years away from commercial readiness. And here's why. If we take a deeper look into Verizon earnings call their CAPEX spending for 2008 was almost the same as their previous year. LTE deployment would require billions of dollars in CAPEX. Considering the current market economy, it is hard to imagine Verizon going with an aggressive rollout. So as per Verizon CTO Dick Lynch, we should see LTE rollout in 2011. But I believe it will be more of phased approach with all markets covered by 2013. We should see other carriers also follow the same approach.
LTE boasts of download speeds of about 34Mbps (I would assume it will be 10-15Mbps theoretically). LTE will be a platform for all data centric applications requiring higher speeds in the order of 8-10 Mbps. VoIP should also take off. But the phased approach in LTE deployment would require current 3G networks to support higher speeds. EVDO RevA can only support upto 3Mbps, though theoretically its must less than that (700Kbps-1Mbps).
Here is where EVDO RevB steps in for the 3GPP2 camp. For carriers like Verizon, it should only be a software upgrade not requiring much additional spending. 3 carrier RevB can support downlink speeds of upto 9Mbps and uplink speeds of upto 5Mbps. Qualcomm also announced availability of MSM8960 chipset supporting both LTE and EVDO RevB at MWC 2009. So if carriers decide to roll out RevB along with phased out LTE approach, they should see availability of dual mode handsets supporting LTE/RevB handovers with seemless mobility. This would enable applications designed for LTE getting traction soon resulting in rich user experience. Nobody wants their video conferencing session being closed just because they moved from LTE to EVDO network.
Now RevB makes even more sense to Sprint. The beleaguered carrier just lost 1.4M subs in Q4. WiMax even though being a good technology with time to market advantage, its facing the same destiny as EVDV. So even if WiMax is prevalent, I see it more of a backhaul solution than 802.16e. Moreover its now Clearwire's headache. So now Sprint is left with its EVDO RevA network. Their rival Verizon is moving to LTE. So they need a network that can compete with Verizon in terms of speed, supporing data hogging applications. And year on year there has been consistent increase in data usage. Cisco even predicts data usage going up with video services accounting for 64% of mobile traffic by 2013. Surely if the carriers are to support around 5GB demand for data per month they ought to have network that can meet the demand. With EVDO RevA I do not see that happening. Also the incremental cost of deploying RevB will be much less.
I think both Verizon and Sprint should go for RevB in order to get maximum out of their current 3G networks before we see 4G taking over. 3GPP camp will also do the same with Tmobile, AT&T upgrading their networks to HSPA + coupled with phased LTE rollout.
Subscribe to:
Posts (Atom)
