1. Enhancing Bluetooth TCP Throughput via Packet Type Adaptation Ling-Jyh Chen, Rohit Kapoor, M. Y. Sanadidi, Mario Gerla Dept. of Computer Science, UCLA

2. ICC 20042 Outline of the Talk The problem: Wireless interference and bit errors severely affect TCP efficiency. The opportunity: Bluetooth offers multiple packet type options with different FEC and packet lengths. Moreover, the link layer API provides link error quality information. Opportunity for cross-layer adaptation. Key idea: dynamically select packet type based on measured link quality. The results: we show that the “Adaptive Packet Type” approach in Bluetooth can effectively enhance TCP performance.

3. ICC 20043 Who still remembers Bluetooth? Application Examples Automatic synchronization of calendars, address books, business cards Proximity operation (camera to cellphone, etc) Personal Area Network Designed for “cable” replacement

4. ICC 20044 Bluetooth Physical Link Point to point link  master - slave relationship  radios can function as masters or slaves ms s s m s Piconet  Master can connect to 7 slaves  Each piconet has max capacity =1 Mbps  Frequency hopping pattern is determined by the master

5. ICC 20045 Piconet Topology Master Active Slave Parked Slave Standby Page - scan protocol Polling

6. ICC 20046 Piconet MAC Protocol : TDM/Polling m s1s1 s2s2 625 µsec f1 f2 f3 f4 1600 hops/sec f5 f6

7. ICC 20047 Multi Slot Packets m s1s1 s2s2 625 µsec f1 f4 f5 f6 Data rate depends on type of packet

8. ICC 20048 Data Packet Types DM1 DM3 DM5 DH1 DH3 DH5 2/3 FEC No FEC Symmetric Asymmetric 108.8 258.1387.2 54.4 286.7477.8 36.3 Symmetric Asymmetric 172.8 390.4585.686.4 433.9723.257.6

9. ICC 20049 Scatternet

10. 10 Bluetooth packet types DH: Stop and Wait ARQ DM: ARQ as well as 2/3 FEC codes to correct single bit errors FEC coding scheme:  (15, 10) Hamming code,  each block of 10 information bits is encoded into a 15 bit codeword  can correcting a single bit error in each block.

11. ICC 200411 Throughput Analysis DH mode: (ARQ) PER: 1 hop Throughput: 2 hop Throughput: DM mode: (ARQ+FEC) PER: 1 hop Throughput: 2 hop Throughput: P: Packet Error Rate, B: Bit Error Rate, S: Packet Size, T: Max Throughput

12. ICC 200412 PER vs BER

13. ICC 200413 ModeBER range DH5<0.0001529 DM5>0.0001529, and <0.0060795 DM3>0.0060695, and <0.0157813 DM1>0.0157813 Bluetooth Throughput

14. ICC 200414 Proposed Approach Adaptive Packet Type (APT):  In BT specs, the function call, Get_Link_Quality, returns the Quality of the specified Link.  We read the returned link Quality Value, and adapt packet type so as to optimize throughput.

15. ICC 200415 Simulation 1: Fixed BER Time: 600 seconds TCP Packet Size: 500 bytes Buffer Size: 9000 bytes

16. ICC 200416 Simulation 2: Varying BER Time: 600 seconds TCP Packet Size: 500 bytes Buffer Size: 9000 bytes BER: changes between 0.0001 and 0.0005 every 1 second

17. ICC 200417 Simulation 3: Measured BER Traces 802.11 interference experiments using CSR chipset CSR provides LQ vs BER conversion tables : If BER (Bit Error Rate) = 0, LQ (Link Quality) = 255; perfect channel. If BER <= 40/40000, LQ = 255 – BER * 40000. If 40/40000 < BER <= 4000/40000, LQ = 215 – ((BER / 32) * 40000). If 4000/40000 < BER <= 40000/40000, LQ = 105 – ((BER / 256) * 40000). Simulation: Time: 600 seconds TCP Packet Size: 500 bytes Buffer Size: 9000 bytes BER: using the BER trace

18. ICC 200418 Simulation 3: measured BER trace

19. ICC 200419 Conclusions In Bluetooth, TCP throughput collapses with BER above 0.03% (eg, BER caused by near 802.11 interference) APT (Adaptive Packet Type) approach can restore TCP throughput to acceptable values for much higher BER (we tested up to.3%) APT technique can be applied to any wireless link with packet length and FEC options, and with link quality (ie BER) feedback. Further work on BT crosslayer optimization will include:  Adaptive optimization of number of retransmissions (for a mix of TCP and real time traffic)  Interleaved FEC over multiple frames

20. ICC 200420 T h a n k you