LTE: A Latent Problem?

Analysts are divided about how latency might affect the performance of Long Term Evolution (LTE) networks in the U.S., following the publication of recent network test results. (See AT&T LTE Test-Drive Results Revealed.)

Signals Research conducted a drive test of AT&T Inc. (NYSE: T)'s LTE network in Houston, using measurement gear from Accuver, which provided the first performance metrics of the newest LTE network in the U.S. The test measured downlink and uplink speed, latency and handover times between LTE and 3G networks, and Light Reading Mobile reported the high-level results here when they were released.

According to Mike Thelander, CEO and founder of Signals Research, the latency times on AT&T's LTE network were disappointing but in line with other LTE networks in North America. The average latency was 49 milliseconds with a minimum of 40 milliseconds, which was "higher [worse] than we were expecting based on the capabilities of LTE and our experiences with testing networks in Europe," noted Signals Research in its report.

By comparison, on LTE networks in Europe, the consultancy has found average latency times of 15 milliseconds to an external server. And back in Houston, Signals also recorded a minimum round trip latency of 27 milliseconds on a commercial 3G network, which was not identified.

But a latency measurement of less than 50 milliseconds is not something to be disappointed about, according to Gabriel Brown, senior analyst at Heavy Reading. "It's a bit like saying, I'm disappointed in this BMW because it's not a Rolls-Royce."

Brown added, "It’s a new technology and there is scope for optimization -- for example, of the EPC architecture and underlying IP network -- that could help drive this figure lower in future."

And that's something the analysts do agree on -- that the LTE networks have yet to be optimized fully. And there are myriad network design and architecture choices that affect latency.

According to Thelander, the latency issue on U.S. networks could be related to how the carriers have integrated their wireline and wireless networks, for example. "They could do a better job of leveraging their wireless assets and do a better job with how they connect to the Internet," he said.

Why this matters
Bandwidth and latency are the two key elements that determine the perceived performance of a network connection. Latency refers to the time it takes to process data on a network, and excessive latency can lead to wobbly connections and delays.

Network latency, perhaps more so than average downlink speeds alone, can affect users' experience, especially for real-time services like video calling, VoIP and even gaming applications.

For more
Check out these other early results from new LTE networks in the U.S. and Europe to see how they compare to AT&T:

— Michelle Donegan, European Editor, Light Reading Mobile

Page 1 / 2   >   >>
Michelle Donegan 12/5/2012 | 4:51:33 PM
re: LTE: A Latent Problem?

Perhaps this adds a bit of context to the results, as I understand them: Signals measured latency to the same server on ATT's LTE network, and, for comparison, on Verizon FiOS as well as on a 3G network. On ATT, the average was 49 ms; on FiOS the average was 14 ms; and on the 3G network it was as low as 27 ms. 

"Low latency" is one of the key features of the LTE spec, but it looks like there's some debate out there about what "low" is. 

Talking about all this in milliseconds is one thing, but I'd like to know when these differences are noticeable? Can users notice a difference in their service between say, 20 ms latency and 49 ms?

tankbert 12/5/2012 | 4:51:33 PM
re: LTE: A Latent Problem?

Comparing FiOS to a wireless connection seems a bit silly. Also 14 ms on GPON seems to be quite high, unless the sophistication of their testing is simply "ping google" from various connections. Latency to the first hop is more important to determine purely the access technology's latency, and end-to-end latency is more of a characterization of load/oversubscription during peak hours in the service provider's backhaul/core network. It's possible the first hop on some of these networks isn't pingable however, so without some service provider assistance it may very well be as unsophisticated as pinging some known fixed server on the internet.

I also don't believe the 27 ms 3G measurement was taken during any time of significant load on the network, as once you're on a populated 3G cell round trip times go up quite quickly from my experience/measurements on various 3G providers in the US.

If the radio resource management in LTE is far superior to 3G, the ability to maintain lower latencies for many more users during load is much more impressive of an accomplishment than being concerned about the lowest possible number that can be reached during ideal network/radio conditions. If 20 people on the same cell site can have 50 ms latency simultaneously with bandwidth porportionally shared between them, that's a much better user experience than either 1 person being able to have 30 ms on 3G but 20 people having 150 ms.

Can a user discern between 20 and 49? Not for a 2-way voice or video call for sure, but maybe a stock broker that wants to do high speed trading in the park :)

tankbert 12/5/2012 | 4:51:33 PM
re: LTE: A Latent Problem?

Considering latency between a subscriber's home and their first hop (interleaved DSL/VDSL with a DSLAM or DOCSIS with a CMTS) can be 30-40+ ms, I find the disappointment in the results somewhat absurd. Signals Research needs to watch this:





OpEd 12/5/2012 | 4:51:32 PM
re: LTE: A Latent Problem?

If you don't like the numbers you are seeing now, wait until there are at least a few other users on the network, and some tens of thousands of small cells connected back to the LTE core. I think all you will need is a bog standard stopwatch to measure the latency then!

We will see some LTE functions shifting their locations in order to cope, IMHO.


Mike Thelander 12/5/2012 | 4:51:32 PM
re: LTE: A Latent Problem?

It is important to remember that a number of popular websites require literally dozens of interactions between the mobile device and the host site to load the page.  In this context, a difference of 20ms RTT is additive, and it can equate to a couple of seconds difference with websites that require a lot of HTTP interaction - something even the casual user would notice.

LTE was designed to support sub-20ms latency and I've seen closer to 12ms in commercial networks going to a server sitting outside of the network.  If the latency is much higher than the theoretical/achievable limits and on-par or even worse than what can be achieved with HSPA+, then I think there is definitely room for improvement.

If operators can make these improvements and if vendors can address current scheduler/performance deficiencies that exist then it could easily be the case that the performance on AT&T's and VZW's LTE networks will be better in several months time, despite the higher number of subscribers using the networks.

paolo.franzoi 12/5/2012 | 4:51:31 PM
re: LTE: A Latent Problem?


AT&T last time I checked had two switching nodes in the US...if you are 1000 miles from one that is a latency of like 10msec at the speed of light.  Which you can't achieve,




Mike Thelander 12/5/2012 | 4:51:31 PM
re: LTE: A Latent Problem?

Not sure if you mean mobile subscriber station or managed security services.  In any event we measured the RTT from the mobile device to a variety of servers/web sites, including private FTP servers, Yahoo, CNN, iTunes, Amazon, etc.  Latency for the latter was based on the ACK Synch times during the initial handshake between the device and the site/server and not a typical "ping test," which these sites generally block.

I wouldn't expect LTE to achieve 15ms or even 20ms to those sites - the FiOS value was only provided as a benchmark and to provide some perspective on what the incremental latency could be due to transport delay, routing, etc., that falls outside of the operator's network. 

On a relative basis LTE latency should be lower than all other technologies/networks in the US (EV-DO Rev A, Mobile WiMAX, HSPA+).  However, that is not necessarily the case, in particular when the comparison is with T-Mobile's HSPA+/DC-HSDPA network.  In a few extreme cases it is only marginally lower than EV-DO Rev A, but generally that is not the case.

paolo.franzoi 12/5/2012 | 4:51:31 PM
re: LTE: A Latent Problem?


Have you guys calculated the time to get the round trip back to the MSS?  Before you talk too much, you might want to subtract that out before you get over any access architecture.  European carriers tend to have a lot more of them and customers tend to be a lot closer to them.




EyeOfTheNeedle 12/5/2012 | 4:51:30 PM
re: LTE: A Latent Problem?

Consider that latency measured from the UE to a Web server involves multiple segments: LTE, EPC and IP network beyond the PDN-GW. This observed high latency may most likely be a result of an EPC transport bottleneck on S1-U/S5 as opposed to IP network bottlenecks beyond the EPC.

In order to properly characterize the LTE/EPC latency compared to UMTS it seems more appropriate to measure latency and jitter from the UE to first router beyond the PDN-GW.



Mike Thelander 12/5/2012 | 4:51:30 PM
re: LTE: A Latent Problem?

I agree with your suggestion and especially your prognosis.  In this case the objective of the study was to look at underlying performance KPIs (modulation scheme, scheduling, MIMO RI, UE transmit power, etc) and how they vary as a function of other parameters as well as how they vary by technology and by vendor implementation.  Further, we wanted to look at what operators are doing or not doing to optimize their networks to deliver a compelling user experience.

The external network latency measurements that were cited in the article served as a precursor to an analysis of the real world user experience as defined by email synch times, web page download times, Netflix/YouTube video playback, etc.  In that regard the real interest should be on the end-to-end latency since that is what the application must deal with when serving the user.  As you might expect when the end-to-end latency on the LTE network(s) is no better than it is on the HSPA+/DC-HSDPA network then the benefits of using the LTE network are minimal, at best.  It isn't necesarily the fault of the technology, but it is something that the operator and/or its vendors need to address.

Page 1 / 2   >   >>
Sign In