lordsutch

New update: I added the ability to show the physical cell ID (PCI) for any additional LTE sites that are detected, along with the RSRP for the sites. Basically this lets you "peek" in the radio and see if there are any other sites your phone may hand off to. If the phone isn't currently on LTE, it may also show why (i.e. weak signals on all the sites it can find). So far the only devices I know it works on are the Evo 4G LTE and HTC One using a patched Cyanogenmod 10.2 (so, basically, my device). The patches should be going into CM 10.2 proper soon. This may also work on other devices that support LTE (such as the Google Play Edition S4 and One for GSM), although probably not any other CDMA devices. If you do try it with a recent CM 10.2 nightly, beware there is a small buglet in the LTE signal strength reporting of the RIL (for reasons that I don't understand, it's reporting dBm*10 rather than dBm in the place that it stuffs the LTE cell information, even though Android specifies dBm); a patch is in the review process to fix this too, but it may need some tweaking to satisfy the reviewers before it can go in. Linky: https://www.dropbox.com/s/6qpa3tbs31zr5dg/SignalDetector.apk

Alas, that doesn't seem to solve the problem on my device. I must have a blob missing or something...

Another interesting discovery with the 4.3 HTC RIL: if you dump out all the CellInfoLte records, you'll see the PCI, RSSP, and RSSQ for other sites the phone can also see. The site the phone is connected to will be the one with isRegistered()=true and usually will have the GCI and TAC reported as well. I may see about adding that info to my app mostly for my own amusement, but others might find it interesting too. For example: D/CdmaSST ( 930): [CdmaLteSST] SST.getAllCellInfo(): X size=4 list=[CellInfoLte:{mRegistered=YES mTimeStampType=oem_ril mTimeStamp=6598001508117ns CellIdentityLte:{ mMcc=310 mMnc=120 mCi=61107458 mPci=318 mTac=7466} CellSignalStrengthLte: ss=16 rsrp=-1023 rsrq=-66 rssnr=2147483647 cqi=2147483647 ta=2147483647}, CellInfoLte:{mRegistered=NO mTimeStampType=oem_ril mTimeStamp=6598001508117ns CellIdentityLte:{ mMcc=2147483647 mMnc=2147483647 mCi=2147483647 mPci=487 mTac=2147483647} CellSignalStrengthLte: ss=13 rsrp=-1108 rsrq=-148 rssnr=2147483647 cqi=2147483647 ta=0}, CellInfoLte:{mRegistered=NO mTimeStampType=oem_ril mTimeStamp=6598001508117ns CellIdentityLte:{ mMcc=2147483647 mMnc=2147483647 mCi=2147483647 mPci=460 mTac=2147483647} CellSignalStrengthLte: ss=7 rsrp=-1259 rsrq=-160 rssnr=2147483647 cqi=2147483647 ta=0}, CellInfoLte:{mRegistered=NO mTimeStampType=oem_ril mTimeStamp=6598001508117ns CellIdentityLte:{ mMcc=2147483647 mMnc=2147483647 mCi=2147483647 mPci=39 mTac=2147483647} CellSignalStrengthLte: ss=13 rsrp=-1796 rsrq=-800 rssnr=2147483647 cqi=2147483647 ta=0}] Here we can see my connected GCI (3A46D02) and its PCI (318) first, followed by 3 other sites my phone is thinking of handing off to: PCIs 487, 460, and 39. 487 is presumably another sector of 3A46D (I think it's 3A46D03) by the Ericsson offset-of-169 rule, while 460 and 39 are PCIs from distinct sites that are clearly further away (-125.9 dBm and -179.6 dBm, respectively) that I haven't definitely identified yet. Someone really clever could probably do LTE geolocation based on having info on the signal strength of 3-4 sites.

I'll try that in my next build. Thanks!

Hopefully this one is small and unintrusive enough that it'll go in in the next day or two. Now to figure out why deck can get a GPS lock on jewel (Evo) but I can't with these blobs... By the way, this patch doesn't fix everything; for some reason I won't have time to delve into for a day or so, apps that listen for CELL_INFO events don't get the updates in the callback even though apps that request an update with getAllCellInfo() do. My cheesy app did the former until today, while both Sensorly and SignalCheck do the latter. I'm not sure it's worth fixing, or even if it can be fixed without deep magic that will get rejected by CM.

Hopefully that means they're going to finally get around to the last few Houston County sites that need tower work (Houston Road at Elberta, Dunbar, the Pleasant Hill water tower, and Feagin Mill Middle School) this week. EDIT: RRUs are up on the site behind the Macon Mall (across from Steak n Shake / Chuck-e-Cheese). Progress marches on. Sent from my Nexus 7 using Tapatalk 4

Cool. I mapped some on the newly live sites in Barnesville and Perry over the weekend. Also, the site in Macon by Coliseum Northside appears to be live, but with authentication off (in what people are calling "donut mode").