Alexandre, I posed this question "How much of Brazil has wireless data?" to a friend of mine who has done some work in Brazil in the wireless telecom area, here is his reply: *** Beginning of message. *** Not a simple question to answer.. First, let me say that the b-band regions mentioned below are only testi= ng CDPD & have not deployed it throughout their licensed area. The country of Brazil was divided into 10 regions for the purpose of issu= ing the b-band cellular license. Our customer operates a network in Region 1 & 10. Their network uses Nort= el infrastructure, where CDPD is easily implemented. The voice protocol in those regions is strictly TDMA. No AMPS, & no analog control channels. I know there is some interest in at least 2 other regions, but they have Ericsson infrastructure, & it is much harder to implement CDPD in their network. I think about 5 of the 10 b-band regions chose Nortel infrastructure. So, there would be 5 where it can be done easily, & the others would be harde= r. FYI....here is the areas where CDPD is being tested or there is an intere= st in it. 1) Region 1 - Sao Paulo Metropolitan area - Nortel b-band infrastructure = - TDMA - Testing CDPD 2) Region 10 - The states of Piau=ED, Cear=E1, Rio Grande do Norte, Para=ED= ba, Pernambuco and Alagoas - Nortel b-band infrastructure - TDMA - Testing CDPD 3) Region 2 - Sao Paulo state - Ericsson Infrastructure - TDMA - Limited Testing of CDPD 4) Region 7 - Federal District (DF), Goi=E1s, Mato Grosso, Mato Grosso do= Sul, Tocantins, Rond=F4nia and Acre states - Nortel b-band - TDMA - Testing= CDPD 5) Region 3 - Rio de Janeiro and Esp=EDrito Santo states - Ericsson Infrastructure - TDMA - Limited Testing of CDPD I do not recall the details of other b-band regions. The 5 regions above covers more than half of the total population of Brazil. The Amazon regio= n (not listed above) had to be auctioned more than once because nobody want= ed it. The area is very large & mostly unpopulated. The a-band is a mix of AMPS only, CDMA & AMPS or TDMA & AMPS. Their area = of coverage is also different than the b-band networks. GSM is the newest networks in Brazil, therefore it has the smallest cover= age areas. Brazil was divided into 3 regions for the purpose of the GSM licen= se. It's a tough call. I wish the guy luck. If he needs data coverage of all = of Brazil for 8KB of data, may I suggest that he look into a commercial HF frequency. If he can narrow his service area, perhaps CDPD, TDMA Circuit Switched Data, or CDMA Circuit Switched Data may be the solution. Also, Nextel is in Brazil & they have some kind of data service. I do not know details of their coverage area. Hope this helps. XXXXXXXX *** End of message. *** Jim ----- Original Message ----- From: "Alexandre Guimar=E3es" To: Sent: Sunday, March 03, 2002 7:13 PM Subject: Re: [PIC]: FSK modem for using with cellphones. > Hi > > > > > DTMF would be too slow, even for transfering just 8k of data a= s I > > need. > > > The redundancy is ok, the files are very small. I can transmit the > blocks > > 3 > > > or 4 times and use small blocks to have a better chance of getting > trough. > > > > In which case 300 baud and forward error correcting (eg Reed Solomon) with > > small packet sizes and a simple ack/nack packet protocol may be a bet= ter > > choice. > > 300 baud maybe a better choice but my usual experience on the old d= ays > of 1200 half duplex modem and 300 baud full duplex modems is that even when > the line is very, very bad the 1200 is much faster. I guess it is becau= se it > is faster when getting a block trough. Less time to get noise into the > signal. We developed a protocol that was adaptive on the block size and went > all the way from 16 bytes per block up to 4096. In very good connection= s the > protocol went from 256 to 4096 and settled at 1024 as the best troughpu= t > point. At very bad connections usually the best troughput was at 32 byt= es ! > When we changed the baud rate this changed completely. We could get 120= 0 > baud connections that were usable where 300 baud connections did not tranfer > anything ! It all depends on how long the noise last and what burst lenghts > it has. > > I will have to experiment with that also. The option of using an > external PLL decoder is not exactly what I was hoping to use but maybe = the > best option really. It should be less disturbed by noise than a > microprocessor decoder. Does anyone knows about good articles online dealing > with software PLL's ? > > Best regards, > Alexandre Guimaraes > > -- > http://www.piclist.com hint: PICList Posts must start with ONE topic: > [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads > > -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email listserv@mitvma.mit.edu with SET PICList DIGEST in the body