×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Hand-Off, Reconnect To Verizon LTE Can Take 2 Minutes

Soulskill posted more than 3 years ago | from the did-you-try-blowing-in-it dept.

Wireless Networking 53

CWmike writes "Verizon Wireless admits that laptop users with USB modems on its new high-speed LTE network may experience up to a two minute delay when switching over from a 3G coverage zone. 'Hand-offs can take up to a couple minutes, but that was expected and a fix is in the works,' Verizon spokesman Jeffrey Nelson told Computerworld. Also, to get back on LTE once already having been in an LTE zone, one tester said it was necessary to unplug the modem and plug it back in again. Nelson said that was not necessary, although he did say it could take up to two minutes to reconnect to LTE. Nelson said Verizon was working with a modem maker on the hand-off problem, but didn't say which maker. Business Insider said the hand-off problem occurred with an LG model. 'We're working with the modem maker for quick update, but no ETA yet.' Nelson said. A Mac OS version was also said to be close."

cancel ×
This is a preview of your comment

No Comment Title Entered

Anonymous Coward 1 minute ago

No Comment Entered

53 comments

"...but that was expected..." (1)

Anonymous Coward | more than 3 years ago | (#34542530)

By whom?

Re:"...but that was expected..." (0)

Anonymous Coward | more than 3 years ago | (#34542656)

Logically, by the engineers who built and tested them, noticing some problems with them.

Re:"...but that was expected..." (3, Insightful)

suomynonAyletamitlU (1618513) | more than 3 years ago | (#34543068)

Logically, by the engineers who built and tested them

...and who utterly failed to convince marketing and sales that it could even possibly be an issue, let alone an embarrassing, glaringly obvious one.

"What are you talking about? The calendar says we ship today. Are you really trying to second-guess the calendar?"

Re:"...but that was expected..." (1)

alen (225700) | more than 3 years ago | (#34546310)

it's not a showstopper bug so the product can ship and be fixed later. it's almost 2011, get with the program. even TV's have service packs these days

Re:"...but that was expected..." (0)

Anonymous Coward | more than 3 years ago | (#34556634)

Thanks, you just proved my point. They were rushed to release, so they did, but they were aware of the bug and would spend time after the release to work on them.

Tick tock. (1)

American AC in Paris (230456) | more than 3 years ago | (#34542554)

You might want to get this sorted out by Christmas, folks.

Just in case, y'know, you release a high-demand 4G mobile device anytime around then.

Version 1.0 (1)

Enry (630) | more than 3 years ago | (#34542590)

This is news to anyone who gets the first iteration of a new product?

Re:Version 1.0 (1)

arivanov (12034) | more than 3 years ago | (#34544352)

This is news to anyone who gets the first iteration of a new 3GPPproduct?

Corrected this for ya.

It is not entirely unexpected when standartisation madness runs supreme. FFS, the world has long become Ethernet based. It takes less effort to originate IP from the terminal and NAT or route IPv6 than to translate to PPP and/or something similarly obsolete and Layer 2 in something that pretends to be a "modem". WTF is a "modem" and why it is a "modem"

If the "modem" is actually doing IP what is it running on is utterly irrelevant. It also makes the entire hand-over malarkey similarly irrelevant because in that case you can the mobile IPv6 stack as used in LTE into the BLOODY DONGLE so it simply continues onto 3G as an "unlicensed mobile access" instead of true handover. In fact it can jump to WiFi, WiMax, RFC1149 pigeons as needed. As there is _NO_ true handover on the network side and network does virtually bugger all it takes seconds, not minutes. Similarly, the network can use custom APNs for 3G and GPRS for these and turn off most of the mobility support on these as it is now the _TERMINAL_ doing it.

Most importantly - implementing all of this in the dongle takes _LESS_ resource than implementing all the fine points of 3G handover (especially the soft handover part).

Sigh... Higher brain functions and 3GPP architecture clearly do not mix well... It is a classic example of Braindead Architecture (TM)...

Wireless Modem Problems (1)

Garrett Fox (970174) | more than 3 years ago | (#34542604)

I don't think wireless modem problems in general should surprise anyone. I'm using a Verizon USB device right now, and while it usually works all right, it has random weird misbehavior like claiming to be connected yet being unable to do anything, or claiming an "invalid username and password" when no new user info has been entered. Probably just standard hardware quirkiness.

Re:Wireless Modem Problems (1)

trum4n (982031) | more than 3 years ago | (#34542868)

my 40$/month cricket USB modem NEVER did that. also, my Sprint phone teathers flawlessly. still cheaper than verizon...and without all the suck. was a verizon user for 10 years, then grew up.

Re:Wireless Modem Problems (2)

willzzz (701172) | more than 3 years ago | (#34543574)

Try getting 35+Mbps down and 5-8Mbps up on Sprint when their WiMAX capacity is under-engineered on average. VZW has all their LTE cell sites on FIBER with a shit-load of capacity.

Glad I didn't sign up... (2)

suprcvic (684521) | more than 3 years ago | (#34542608)

My Verizon 3G service that I use for my business (no wired DSL or Cable available) was out the other day and while I was on the phone with Verizon figuring it out, they were trying to upsell me on this. With a 5GB data limit, I wasn't about to since I routinely go over it on the 3G. I just switched back to Clear who swears they got THEIR reliability issues sorted out in my area.

Re:Glad I didn't sign up... (2)

Pootie Tang (414915) | more than 3 years ago | (#34542776)

Clear is a bunch of liars:

http://i.imgur.com/jTnpJ.png [imgur.com]

That image isn't mine, but I got the same "not junk mail" from them and it is my only impression of them so far. I wasn't impressed by their mailing.

Re:Glad I didn't sign up... (1)

willzzz (701172) | more than 3 years ago | (#34543570)

What? You're EVDO should improve and maxing it out. Areas where the cell towers have LTE deployed have fiber back-haul deployed on the RAN with GigE or higher. LTE has more capacity per sector than cable DOCSIS-2.0 in Verizon's configuration and the same as DOCSIS-3.0 if VZW doubled the spectrum like the European carriers have (e.g. TeliaSonera). You will get 5-35+Mbps and 5-8Mbps up and pings of 30ms within the RAN and 45-100ms to the internet if it's peered/transit is good. Seriously yes there's bandwidth limits. But I'd rather have bandwidth limits and good pings and performance (cable modem/fiber link) than shitty Clear and under-engineered network where it's slow as molasses since they don't have enough backhaul everywhere.

On the plus side (3, Funny)

Anonymous Coward | more than 3 years ago | (#34542630)

This means it now takes 2 minutes longer to exhaust your monthly quota.

This is news? (2)

mr100percent (57156) | more than 3 years ago | (#34542632)

Designers have to strike a balance; if it polls the towers too often and tries to connect via 4G, it would drain the battery and possibly slow down the connection. Switching from a 2G to 3G network on AT&T is similiar, although usually its about 30 seconds

Apparently (was:This is news?) (1)

Zero__Kelvin (151819) | more than 3 years ago | (#34542810)

This doesn't have anything to do with the frequency of polling the towers. Once connection is lost, it is a safe bet that it is time to scan for towers immediately.

Re:Apparently (was:This is news?) (2)

mr100percent (57156) | more than 3 years ago | (#34543864)

Read the summary or article again, it is talking about switching from 3G to LTE (4G?), not from No Service to LTE. Typically it doesn't poll as often AFAIK.

Re:This is news? (0)

Anonymous Coward | more than 3 years ago | (#34544772)

There is no reason why handover from 2G to 3G or the other way should that that long. My smartphone jumps from HSPA to Edge in less than a second.

For some reason equipment manufacturers didn't prepare and make modems that can handle 2G/3G/4G and had them ready for the launch of commercial LTE networks. That is the reason why this type of problem is occurring right now (Telia in Sweden required you to swap USB dongles when you wanted to change from LTE to 2G/3G).

Grammar (0)

Anonymous Coward | more than 3 years ago | (#34542640)

Hand-offs can take up to a couple minutes

A couple OF minutes.

You wouldn't say "A glass water" would you?

Re:Grammar (1)

the_bard17 (626642) | more than 3 years ago | (#34542858)

No, but I might say that I have two balls. I would not say that I have a couple of balls.

I'd also say that I have three books on my shelf. I would not say I have a few of books on my shelf.

Re:Grammar (1)

vigmeister (1112659) | more than 3 years ago | (#34542884)

If your going too be pedantic, better you are in right!

http://dictionary.reference.com/browse/couple [reference.com]
http://en.wiktionary.org/wiki/couple [wiktionary.org]

Cheers!

Re:Grammar (1)

Anonymous Coward | more than 3 years ago | (#34542978)

your ? too? ... epic fail?
- woosh

Re:Grammar (1)

MacGyver2210 (1053110) | more than 3 years ago | (#34543000)

"–verb (used with object)
10. to unite in marriage or in sexual union. "

Now we're talking!

Two minutes in heaven is better than one minute in heaven.

Re:Grammar (0)

Anonymous Coward | more than 3 years ago | (#34548564)

+1 Flight of the Conchords

LTE seems like a rip off (3, Insightful)

cervo (626632) | more than 3 years ago | (#34542712)

first off, this piss poor service that was just advertised. Two minutes to reconnect?? Sounds like it isn't ready for prime time...

Next, look at the per use charges. I'd rather have a slower connection with much hire quotas (or no quota) than a super fast connection with a tiny quota. Good luck to anyone who wants to watch videos.

As it is cell service is a huge rip off, and LTE is even more of a rip off than 3G. But what do you expect from providers who charge a fortune for delivering simple text messages and the rate hasn't gone down as their networks are upgraded...In fact ATT's rate went up (they used to be free to receive, but after the Cingular merger the double charging started....).

Re:LTE seems like a rip off (0)

Anonymous Coward | more than 3 years ago | (#34543018)

If LTE is poor, then we're screwed, because it's the only prospective 4G service planned AFAIK. WiMAX and HSPA+ can go to hell.

Re:LTE seems like a rip off (1)

willzzz (701172) | more than 3 years ago | (#34543558)

NO only switching between EVDO rA and LTE. They are COMPLETELY TWO SEPERATE technologies on the BACKEND with different authentication though supposedly with ALU and Ericsson VZW integrated the network cores with the new EPC (Evolved Packet Core)... This is a DRIVER PROBLEM with the LG VL600. The UML290 by Pantech uses a newer Qualcomm chipset and DOESNT HAVE THIS PROBLEM. More of an OEM problem than a network problem... As to the costs... Network building COSTS MONEY. EVERY SINGLE CELL SITE (99.99%) is back-hauled by fiber. Or Microwave of similar capacity if they can't get the fiber at the moment. You want high bandwidth? Pay for the network buildout and fiber backhaul! VZW is doing it right for once.

Re:LTE seems like a rip off (1)

tautog (46259) | more than 3 years ago | (#34545220)

Enough, Verizon shill. VZW has so oversold their bandwidth since the introduction of the original "Droid" that their "3G" (more accurately, 2.5G) network is all but unusable. Their highspeed service has always sucked (been with them for nearly a decade with smart phones and usb dongles) and the only reason I've tolerated it was because their voice service worked in all the backwater shitholes I have had to frequent. Now that I'm no longer relegated to traveling the gravel roads and trails of the rural MW USA, I'm planning to switch to a different network (for instance, T-Mobile provides better voice service to my fairly rural home). The extended benefit will be not having to deal with their shitty customer service and horrible pricing structure.

Just because the backhaul from the tower is fiber doesn't mean that the backbone or infrastructure can tolerate the load. You can be assured that VZW will half-ass this deployment, as they do with everything else.

Re:LTE seems like a rip off (1)

Konshu (1643829) | more than 3 years ago | (#34553276)

I think you would be a great candidate to start a new company that moves away from such poor business models, quick get on kickstarter now!

Similar problems with Sprint 3G/WiMAX.. (2)

faedle (114018) | more than 3 years ago | (#34542964)

While I can't say it takes two minutes, the handoff between the WiMAX network and 3G is very rough.

For starters, since they're literally two separate networks, you will lose any open TCP connections because the IP address changes. If you are in a marginal 4G coverage area, you can have it bounce between the (poor) 4G signal and 3G network. The Overdrive hotspot has settings for fine-tuning how quickly it will try to switch and how often, but most of the time unless I'm in a known-strong 4G coverage area it's better to just lock the device at 3G and not even let it try to swap. Similarly, if you're in a strong 4G area the reverse is true.

Secondarily, I have nothing but problems trying to use any Sprint 4G device in Seattle. Both my hotspot and my Evo 4G phone do weird things (often crashing entirely), but only in that market (I've had no trouble in other markets, especially my home market of Portland). The Evo 4G will actually crash about 20% of the time I switch the 4G on up there.

I think that they really need to work on making the two networks as seamless as possible. I know that's not going to be easy from a technical perspective, but it kinda shows that the technology is not quite ready for prime time.

Re:Similar problems with Sprint 3G/WiMAX.. (1)

Nyder (754090) | more than 3 years ago | (#34543538)

While I can't say it takes two minutes, the handoff between the WiMAX network and 3G is very rough.

For starters, since they're literally two separate networks, you will lose any open TCP connections because the IP address changes. If you are in a marginal 4G coverage area, you can have it bounce between the (poor) 4G signal and 3G network. The Overdrive hotspot has settings for fine-tuning how quickly it will try to switch and how often, but most of the time unless I'm in a known-strong 4G coverage area it's better to just lock the device at 3G and not even let it try to swap. Similarly, if you're in a strong 4G area the reverse is true.

Secondarily, I have nothing but problems trying to use any Sprint 4G device in Seattle. Both my hotspot and my Evo 4G phone do weird things (often crashing entirely), but only in that market (I've had no trouble in other markets, especially my home market of Portland). The Evo 4G will actually crash about 20% of the time I switch the 4G on up there.

I think that they really need to work on making the two networks as seamless as possible. I know that's not going to be easy from a technical perspective, but it kinda shows that the technology is not quite ready for prime time.

Not sure if this will make you feel any better, but I have problems with 3G coverage in Seattle. Well, had problems. Decided to let my cell phone go. No home phone, no cell phone. I win!!!

Re:Similar problems with Sprint 3G/WiMAX.. (1)

Anonymous Coward | more than 3 years ago | (#34544266)

There is no comparison between 3G/WiMAX handover and 3G/LTE handover.

3G & WiMAX are two separate networks and there is no true handover between them, just disconnect/connect.

However, 3G/LTE handover is standardized in 3GPP and you can maintain packet connectivity since you will basically keep your GTP tunnel to the GGSN (the GTP endpoint & packet router). This is very similar to 2G/3G handover, and 2G/LTE handover should be possible as well.

These current issues are due to the infancy of LTE, and due to Verizon being one of the pioneering LTE networks (using the very first dongles). Everything should be ironed out within the coming months.

No crashes with Samsung Epic in Portland (1)

linuxguy (98493) | more than 3 years ago | (#34547138)

Just adding some additional data. I have a Samsung Epic in Portland area. I sometimes use 4G on my phone. I do not believe my phone has ever crashed in the 2 months I have owned it.

No Verizon Crapware! (1)

MacGyver2210 (1053110) | more than 3 years ago | (#34542988)

"A Mac OS version was also said to be close"

I hope you mean the driver. If I have to install some shitty software you monopolizing bloatware manufacturing retards came up with, the deal is completely off the table.

Re:No Verizon Crapware! (1)

willzzz (701172) | more than 3 years ago | (#34543548)

It's not a driver problem, it's a software problem. If you don't want to use VZAccess you don't have to, you just don't get any support. Anyways if you understand the engineering fundamentals and the technical specifics read this: http://community.vzw.com/t5/4G-Discussion/4G-LTE-Data-stick-Mac-Linux-Windows-other-authentication/td-p/347794 [vzw.com] The LTE UML290 card supports standard GSM 3G/4G/LTE stack and works just fine with Apple's generic dialer or the Windows connection thing in Win7 or Linux. The VZAccess support is for the plebs and not developers... You have to run it ONCE in VZAccess though on Windows atm (coming soon for Mac) for authentication and the SIM card programming...

Re:No Verizon Crapware! (1)

StuartHankins (1020819) | more than 3 years ago | (#34546338)

You have to run it ONCE in VZAccess though on Windows atm (coming soon for Mac) for authentication and the SIM card programming

Then the deal is off. I'm not interested in any device that doesn't work out of the box with my Mac.

Re:No Verizon Crapware! (1)

Dog-Cow (21281) | more than 3 years ago | (#34547008)

Verizon is weeping in misery now. Truly. Really.

Ah, who am I kidding? Verizon could not possibly care less that you, personally, will not purchase an LTE dongle for use on their network. It's not as if you were considering it anyway.

Re:No Verizon Crapware! (1)

SpectreBlofeld (886224) | more than 3 years ago | (#34548344)

Then the deal is off. I'm not interested in any device that doesn't work out of the box with my Mac.

You must miss out on a lot of stuff, then.

Redundancy in summary (0)

Anonymous Coward | more than 3 years ago | (#34543062)

A bit redundant redundant summary.

connecting and reconnecting (1)

pz (113803) | more than 3 years ago | (#34543142)

Why would any inter-computer handshaking process require a scale of minutes? Someone is not thinking here.

Processors are running HUNDREDS OF MILLIONS OF OPS PER SECOND ... how can a simple handshake take more than a few milliseconds?

(I've yet to understand why Wifi connections, as another example, aren't almost instantaneous; communication between two computers, even with really generous timeouts should not take on the scale of seconds.)

Re:connecting and reconnecting (1)

Anonymous Coward | more than 3 years ago | (#34543310)

Do you know anything about how they work, or are you just shitting on something you don't understand?

Re:connecting and reconnecting (1)

pz (113803) | more than 3 years ago | (#34545440)

Do you know anything about how they work, or are you just shitting on something you don't understand?

I have implemented similar identification protocols and never had to include a seconds-long timeout which, I strongly suspect, is what drives most of the delay in, say, getting an IP address from a Wifi base station, or from the local cell tower. It takes microseconds to form and send a packet. If it only takes a few hundred milliseconds to fetch a web page from a site that knows what it's doing all of which is going across the net and not the last hop to your device, then why does it take anything close to *seconds* -- let alone the minutes mentioned in the original posting -- to do an initial negotiation?

There are probably other possibilities, but a standard construct that I've seen time and time again in code I've reviewed is like:

START:
broadcast_im_here();
while (less than one second has elapsed) do
      if reply_waiting()
          set_local_id();
          break;
      end;
end;
goto START if no_id_was_allocated();

But the one-second timeout is on the wrong order of magnitude if you look at the distribution of response times. In one system I'm intimately familiar with, the time spend on making connections and the subsequent short communication was far, far too long. Looking at the distribution of time spent in a loop like the stylized one above, most connections were very established very quickly, with a nice tail-off for longer delays. But the longer delays didn't decay fast enough, and thus most of the time was spend on less than 5 percent of the connections. We implemented a better, much shorter timeout before issuing a retry, and the system got much significantly without descending into ID-request storm.

The point is, as programmers, we often think in human-scale times -- one, two, five or even ten seconds is a reasonable timeout for humans. But for machines that operate at a fundamental timescale that's six orders of magnitude faster, it's like trying to order a cup of coffee and giving up after a month because the person behind the counter didn't hear you. Yes, the process still works, but it's horribly inefficient. While I am not familiar with Wifi, 3G, 4G negotiations, etc., in my experience timeout thresholds are often not just a little too long, but vastly too long. When my home AP takes 10 seconds to give my laptop an IP address when there are no other devices around, something's wrong in the protocol design.

Re:connecting and reconnecting (1)

willzzz (701172) | more than 3 years ago | (#34543576)

I don't have the fucking time to descibe the protocol differences of 3GPP and 3GPP2 and protocol fail of LG and their shitty drivers. The UML290 model doesn't have this problem and uses a newer Qualcomm chipset.

More LG Crap (1)

Ada_Rules (260218) | more than 3 years ago | (#34544746)

Really is there anything that LG makes that is not utter and complete crap? Perhaps I just had some uniquely bad luck with the LG products I've purchased but at this point I'd buy almost anything made of conflict diamonds and pureed kitten claws over anything made by LG.

Re:More LG Crap (0)

Anonymous Coward | more than 3 years ago | (#34547670)

I've got an LG toaster and it works splendidly. Though you can't go far wrong with a toaster.

Check for New Comments
Slashdot Account

Need an Account?

Forgot your password?

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>
Sign up for Slashdot Newsletters
Create a Slashdot Account

Loading...