Jump to content

xenadu

S4GRU Premier Sponsor
  • Content Count

    249
  • Joined

  • Last visited

Posts posted by xenadu


  1. In conjunction with my post above, here are some scientific developments that obviate the need for additional spectrum:

     

    The method -- time-domain transmit beamforming -- involves digitally creating a time-domain cancellation signal, coupling it to the radio frequency front end to allow the radio to hear much weaker incoming signals while transmitting strong outgoing signals at the same frequency and same time.

    Today's wireless radios use two separate channels to transmit and receive signals, but full duplex radios transmit signals at the same time in the same frequency band. This can double the efficiency of the spectrum, but the problem is often interference between the transmission and receiving functions on full duplex radios.

     

    http://www.innovationgeneration.com/author.asp?section_id=2807&doc_id=256003&

     

    And algebraic answers to the capacity crunch:

     

    http://www.innovationgeneration.com/author.asp?section_id=2557&doc_id=253598

     

    That woman is an idiot and doesn't understand what she's talking about. TCP isn't licensed from anyone and the technique deals with TCP throttling due to missing packets, where reconstructing the missing packet greatly increases throughput by avoiding both the latency (where all subsequent packets are useless until the retransmit succeeds) and the receive window scaling where TCP tried to adjust to the bandwidth. Basically TCP assumes low packet loss which is often not the case for WiFi links.

     

    Not to mention TCP is not the only protocol used on the Internet.


  2. Goodreader for managing documents, reading files, opening ZIPs, etc

     

    AirVideo for live streaming video from your PC.

     

    Angry Birds Space & Star Wars, Where's my Water, Where's my Perry, ERA and JellyDefense if you like TD games. Bastion is a kickass action game, Drawn: the Painted Tower and its two sequels are Myst-like puzzle adventure games, well worth the price.

     

    Intellicast for weather.

     

    Downcasts beats the built-in podcast app by a mile.

     

    Notability was already mentioned.

     

     

    And of course my app, Storm Sim :)


  3. I believe for future spectrum auctions and mergers the FCC is segmenting spectrum into sub-1Ghz, 1-2.2Ghz, and 2.5+ Ghz with value placed on the spectrum from most to least in that order. It's recognition of the fact that 700Mhz is more valuable than 2.5/2.6Ghz so I don't think the Clearwire spectrum will hurt Sprint at all in bidding for PCS H block.

     


  4. FYI: you aren't supposed to run Ethernet over copper more than 100m/330ft. Fiber would be the only legit way to run a cable that distance and that means conduit.

     

    A directional antenna or some of the various products listed here would be perfectly fine to throw from a home to a barn, etc. 5GHz is almost line-of-sight anyway.

     

    If you want actual coverage over a wide distance you can try illegally boosting the output power of your AP, but your clients won't be boosted and at some point the AP just won't be able to hear them regardless of how high you boost the power... You'd need some custom APs designed to cover a large area and use an outdoor antenna mounted at sufficient height.


  5. Tested this out. There is no "CDMA only" for iPhone 5, so I just checked my IP via Google with LTE on and off.

    When LTE was on, I got 66.87.83.30.

    When I turned it off, I got 66.87.83.10.

    Maybe the iPhone 5 connects solely to eHRPD regardless of LTE switch status? I know I'm on eHRPD because I'm in an active deployment market with multiple LTE sites within a mile of me.

     

    Well if you are connected to the same tower for LTE and 3G then you should be using the same gateway, etc. Even where the older 3G cabinet isn't switched over or upgraded simultaneously, I think they are still moving all backhaul over.

     

    Even if you aren't, there's no reason they can't just be trying to keep your handset on the same IP address to avoid interruptions with streams, VPNs, etc during handoff between cells. Compared to modern router memory, keeping the several million IPs of individual customers within a market individually routed isn't much of a burden. Someone with more knowledge of how those network cores work could probably shed more light on it... Been a long time since I've messed with router configurations and telcom stuff.

    • Like 1

  6. I'd say Sprint and their big brother Verizon has plenty of incentive to drag their feet on VoLTE. The main incentive is that the world has moved more and more away from CDMA2000 and they can still lock-in customers. Verizon doesn't really have to comply with the Open Access rules they agreed to to win 700 Upper C spectrum. That is the big reason that CDMA is the ultimate "hold me back" on consumers. Verizon has at least started unlocking GSM/UMTS on their phones, no doubt due to the prompting of the FCC.

     

    I think technology has finally caught up with them and both Sprint and Verizon have accepted that future devices will use SIM cards with no more ESN/ESID lockouts. There's just no hurry to make it happen as it has costs with zero benefit to them.

     

    VoLTE will be necessary for Sprint as it allows them to eek a bit more out of their spectrum and the equipment is the global standard (and thus will be cheaper), and it turns out the carriers largely lost the handset control battle thanks to Apple (the reason Android has a standard Google Play store without any carrier approvals required - the carriers were desperate to compete with Apple and so was Google). I wonder if some of the data capping policies are a result of trying to make their money that way instead **.

     

     

    ** Standard IP network management is more than sufficient to solve the supposed "data hog" problem by fairly slicing up the available bandwidth during peak usage times.

    If someone is using 50GB of data a month, then you move their traffic to the bottom of the priority queue... At noon, they get 200Kbps speeds. At 3AM? They get 30Mbps of LTE. The equipment is on and the big three cell providers are all Tier-1 backbones so they don't pay for upstream bandwidth or peering at all... It literally costs them nothing for you to download a terabyte over the cell network, as long as it happens when the network is lightly loaded.

    There are a thousand different priority/scheduling/traffic management systems that have been widely deployed for years, none of which require data caps or overage fees. Caps are a pure money play, nothing else. A way to "tax" Google, Netflix, etc.


  7. Well they are spending the money on R&D, it's no surprise their engineers are coming up with some good designs. I think the "antenna gate" issue also drives them to make sure no one can ding their radio performance ever again.

     

    gusherb: I would assume the issue there is just plain old signal absorption... Not much you can do at 1900 MHz. When 800 SMR goes live you should see good improvements for in-building coverage. Around here, my LTE signal at home has gone from 1 bar to 2 bars since the leaves have fallen off the trees. I'm at the edge of an elevation hole and there isn't much anyone can do about it until the 800 SMR LTE gets turned on. Plano is mostly a flat plain but we're near an old creek, one of the few places it isn't.

     

    I know the iPhone doesn't officially list LTE 800 SMR but does list CDMA voice for that frequency. My understanding of the antenna system and the MDM 9615 says there is nothing technical in the hardware that would block this frequency, so I'm really hoping it supports it or can be updated to support it.


  8. I'm pretty sure Sprint is being cautious and slow going after new customers and selling these devices because they need Network Vision to be further along. It will do more long-term damage if they get an even worse reputation for terrible speeds.

     

    I tested the 3G network at my office yesterday... 0.01Kbps download speed.

     

    But when I'm on the third floor I can now barely pickup LTE... Even at such a marginal signal level I was getting 2-3Mbps.

     

    My next iPad will be Sprint cellular; by then LTE should be fully deployed around here and it will be nice to not have to turn on tethering or fuss with that stuff.


  9. I don't want to make you feel old but I've never had a landline since I moved out of my parents house; neither has my wife since I've known her. Except for my brief stay with AT&T (for the iPhone 3GS until the 4S was released) I've been a Sprint customer since high school.

     

    I don't know anyone my age who still has a landline.


  10. Tiered data plans are such a joke, if it were really about managing the network and not increasing profits they would just use QoS. During times of peak load, slow the heaviest users down. During off-peak hours, let the data fly.

     

    The technology doesn't require any new hardware or changes to data plans. If I used 15GB this month and I'm trying to download at 4PM on a Monday, then throttle me down to let the light users through. At midnight when no one is using it, let me go full speed.

     

    Given Sprint's 3G situation I don't think they'll get many takers on a tiered data plan at the same price as AT&T/VZW. Now if I can add an iPad to my monthly plan for 10 bux that's another story.

    • Like 1

  11. My next iPad will probably get Sprint cellular, depending on the prices. I don't know that Sprint has ever done pay as you go, no-contract data plans before.

     

    If you ever need WiFi tethering, the iPad with the screen off just serving up LTE is ridiculous... I've seen people getting 24 hours of tethering with the iPad 3.

    • Like 1

  12. Hah looks like we were right about Clearwire!

     

    This FCC administration seems to be willing to look at spectrum aggregation not as an absolute but as a sum of parts of unequal value. In other words, VZW and AT&T may not be allowed to gobble up nearly all below 1 GHz spectrum, then claim inequity because Sprint-uh-Bank-Crearwire has ~150 MHz of 2.6 GHz spectrum.

     

    If I were the Spectrum Czar that the FCC should have, I would value spectrum according to this formula:

     

    [1000 × bandwidth (GHz)] ÷ [center frequency (GHz)]²

     

    The higher the value, the greater the spectrum applies to the cap...

     

    AJ

     

    You posted exactly what I was going to post.

     

    Any chance that the ordering of spectrum priority changes through all of this? 2.5->1900->800. Many of us were thinking Sprint wouldn't want to pay Clearwire so the priority would be on Sprint frequencies first.

     

    Here is my theory: they aren't buying Clear outright because then that spectrum definitely applies against them in future auctions/mergers, but they need control to prevent anyone else from buying Clear. The goal is to keep them at army's length while they go after MetroPCS, Tmobile, or whoever else... Then bring Clear in-house once those more valuable spectrum deals are done.

     

    I wouldn't be surprised to see Clear sign a hosting deal with Sprint though, to allow them to roll out LTE on 2.6Ghz much faster while still keeping them at arm's length.

     

     

    I'm sure Sprint would love more spectrum in the SMR band but that isn't gonna happen (I always thought they should offer cut-rate service to all existing public safety users in exchange for getting another chunk of SMR... Basically take over the building of a nationwide public safety network, but that depends on the cost).

     

    They need some more PCS spectrum which a MetroPCS or Tmobile deal would get them... If they buy the Dish spectrum that juts right up against the PCS H block so that's a natural fit and nationwide too.

     

    The Clear spectrum is entirely about dense city areas where you don't need the signal to travel far or even to have 100% coverage... You just need to offload enough users to keep average data speeds on your better spectrum high. And deploying a bunch of 20Mhz carriers in the Clear bands does that for you... Plus it's much more likely you can maintain service during large events.


  13. Well next time the DFW schedule map is updated I will take a look at the ILEC/CableCo overlays... If it is backhaul, maybe it is tied to specific providers. I'm not sure if AT&T wholesale picked up contracts around DFW or not, I know they did pickup some. The other ILEC is Verizon so it's possible that Verizon is dragging their feet, or maybe Time Warner cable since their no-compete deal with Verizon. Who knows?

     

    If Sprint just shared more info it would be fine. Knowing the problem is X or crews were pulled to area Y for reason Z would be way better than a lack of information.


  14. Just checking in. I didn't think it was possible for them to get slower but only 2 sites complete in DFW last week, a sharp drop from our previously lightning-fast 8 sites per week.

     

    I really wish I knew what was up or why Sprint can't seem to sustain 200 sites per week total (170), let alone push toward the 500 or so they would have needed to meet their original deadlines. There can be no question now that Network Vision is rolling out far slower than originally planned and I expect ATT and VZW to be 100% done with nationwide LTE before Sprint gets even half done.


  15. They are, but I haven't been finding it consistent. Here in Baltimore, one week we'll have 30 done and the next week only 10 or so.

     

    You're lucky. Here in DFW they only completed 2 sites last week and they've only completed 18% of sites total. On a really good week they'll finish 8 sites. We were originally scheduled to be finished next month (hah) but now it looks like April 2014 (or worse).

     

    I'd love to have 30 sites per week.


  16. This deal is equity to avoid regulatory concerns. It's all about giving Sprint ammo to go after spectrum and smaller companies; they're in an OK position to execute on NV and become profitable but the acquisition wave is coming too soon for them to be ready to take advantage and once that spectrum's gone it ain't coming back. Sprint obviously wants the Clearwire spectrum back in-house and wants (needs) the MetroPCS spectrum.

     

    I predict Sprint will bid for MetroPCS as a result of their strengthened position, or possibly try to merge with T-mobile. I also predict a wave of Sprint acquisitions of regional players, which will pass regulatory muster with the idea that Sprint needs to stay competitive with VZW/ATT. Also look to have them buy out the other Clearwire stakeholders and fold that back in. The result will be a company very well positioned to make the wireless landscape competitive again. The NV platform gives them an easy path to tack on spectrum and the new fiber backbones mean no growing pains. If they can't do one or the other of these deals, look for them to attempt to buy Dish' spectrum or even buy Dish itself. Having learned the lessons of Nextel, look for them to switch over to hosting on their own towers ASAP. They might also finish NV ahead of schedule and look to roll out new towers for expanded coverage.

     

    Why does Softbank want to do this deal? They see the coming wave of mergers/acquisitions. They've been interested in investing in the US for some time and Sprint is relatively cheap right now, has a good platform rolling out, carries the iPhone, and is currently adding customers... things that don't apply to Tmobile. If they can't have Tmobile + Sprint, they'd rather grab Sprint then snatch up the smaller players like US Cellular, Shentel, Leap, etc. Basically they want to buy for the same reason I bought shares when Sprint announced they would carry the iPhone and S dropped to sub-3.00.... because it was a great value and lots of room to profit. For various reasons, Softbank is unlikely to get the same kind of return only investing in Japan.

    • Like 4

  17.  

    We have an iOS app that shows the maps and has the speed test stuff.

    But apple doesn't provide apis for collecting signal strength and, although we have found some hidden APIs, they wouldn't be allowed on the appstore :(

     

    Are you willing to share the source code for the iOS app? Those of us who are developers can certainly compile and install the app; I'm sure some power users would do the same.

     

    You might also try submitting the app, getting the rejection, then apply to the Appeals Board. They overturn rejections regularly... I think you could make the case that a crowd-sourced coverage map is a good thing and access to the data is necessary. Rest of the app had better be polished as s**t though.

    • Like 1

  18. Like I said, it's a trade-off. If you give developers enough rope, they'll hang themselves (or their users), like having an app run in the background continuously instead of making use of the push notification service because doing a thread sleep, fetch url every 30 seconds is way, way easier to write. While either method works, the sleep method consumes a ton of battery checking on things the user may not even care about.

     

    The vast, vast majority of computer/cell phone users don't know or care what permissions, background processes, etc are. They have no idea how to read that information and wouldn't even think to look at it if they did. Half of all adults in the US probably don't know the difference between RAM and Disk/Flash.

     

    So it's just my opinion but as a user, I like knowing that I'll never have to confront a huge permissions dialog or accidentally install malware on my phone. I like knowing I never have to check for a list of running applications or kill a background process, nor wonder if the new app I just installed is eating my battery. That also means I can't run apps like Sensorly**. That's what I meant by "trade-off".

     

    There is one benefit of starting off more restricted: Apple can introduce new APIs or relax restrictions when needed. Android will always be a wild-wild west because taking any of the APIs away or changing them will break tons of existing software. Since devs can just check the boxes to request all permissions in the world, they do because it's easier and almost no one reads or understands that dialog anyway. That's why Microsoft has had such a terrible time with Windows bugs, fragile apps, etc and especially had extremely poor results trying to get apps to stop requiring Administrator access (UAC). Devs and users got so used to it they just assumed it, as a result every random macro, script, or downloaded file runs with the user's full administrator rights. Once the install base is too big the cat's out of the bag and you can't go back and change it without massive headaches.

     

     

     

    ** Of course I can run anything I want on my phone; as a developer I can sign apps, make apps that call into private APIs, etc. Anyone else can do the same if you're willing to part with $100/year for the developer program. You can also use a hackintosh to install apps on test devices, though submitting apps to the store requires a real Mac.

     

    I've maintained all along that Apple should have a "dev light" or "power user" mode buried deep in the options. Charge a small fee ($25?) and have it install a special certificate on the phone or whatever.... that mode would open up and allow most of the stuff jailbreaking gets you today, but in an "at your own risk" way. Normal users would never see it but tech nerds could choose to take responsibility for it if they wanted.

     

    I also think Apple needs a special "System Utility" category that allows root access, full access to the filesystem, etc. Maybe not for the mobile devices, but it is definitely needed for the Mac App Store. Those apps could go through more review, etc but there are a few legitimate cases where such access is needed... but imagine the headlines if Apple blocked some scummy or unknown developer from that category.

     

    They really do strive to be fair and treat everyone the same... they apply the same rules to EA as they do to me. Their App Store appeals board is also quite fair - if you can make the case they can and will overrule the reviewers and have done so on many occasions. My app updates get reviewed within 7 days, just like Rovio or Microsoft, even though one has revenue several orders of magnitude bigger than mine and the other is Apple's competitor. Apple sets the rules but the playing field is level within those rules. As far as I know they don't take promotional payments for app placements either - the search results are what they are, you can't buy your way to the top.

    • Like 3

  19.  

     

    Sounds just like me when I try to navigate around on someone's ipod/iphone. The one button thing drives me nuts as I just want to flip over to another app and not close the one I'm in.

     

    iOS will keep the app suspended in the background unless the app doesn't support it or the system is low on memory.

     

    It's funny... Some of the things you dislike are reasons why I like it. I want someone else to curate the apps for me. I don't wanna have to check for background services to see why my battery life is terrible. Sometimes those are trade offs and I wish I could do XYZ as a developer but on balance I find it to be worth it.


  20. Hey bollar, what apps do you work on? PM me if you can't share publicly.

     

    The iPhone was the first one to do smartphones right, IMHO. Now that I'm in the ecosystem there is no reason to change. Plus as an iOS developer I have an ulterior motive to evangelize the platform.

     

    The support is excellent and I love that the company in control if the iPhone experience is Apple, not the carrier.

    • Like 1
×
×
  • Create New...