doc IEEE 802 15 -TG 2_385 r

doc. : IEEE 802. 15 -TG 2_385 r 0 Project: IEEE 802. 15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [Merged AFH Proposal] Date Submitted: [7 August 01] Source: Address Voice: E-Mail: [Hong Bing Gan, Vitaliy Sapozhnykov, Efstratios Skafidas, Bijan Treister] Company [Bandspeed] [7000 West William Cannon Dr. Suite 2220 Austin, TX 78735] [+1 512. 358. 9000], FAX: [+1 512. 358. 9001], [{h. gan, v. sapozhnykov, e. skafidas, b. treister}@bandspeed. com. au] Re: [AFH Merging of systems] Abstract: [Method of merging of AFH systems] Purpose: [Merge AFH proposal to one unified system while maintaining flexibility] Notice: This document has been prepared to assist the IEEE 802. 15. It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor acknowledges and accepts that this contribution becomes the property of IEEE and may be made publicly available by 802. 15. Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 AFH Merge overview (1) 1) Proposal for AFH merge (Mechanism) – Block diagram – Justification – Comparison and Evaluation 2) Proposal for LMP commands – Use only ‘good’ / ‘bad’ channels in LMP command instead of ‘good’/’bad’/’unused’ Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 AFH Merge overview (2) Logical segmentation: 3. Use remaining ‘bad’ channels in intelligent manner. 23/79 Mode AM_ADDR CLKE 12 -1 / CLKN 12 -1 12 UAP/LAP CLOCK 28 Selection 27 Box fgen Frequency Replacement fadp Grouping / paring Figure 6: AFH System Overview. 1. Bluetooth Hopping Sequence generator Submission 2. Remove as many ‘bad’ channels as required / legally possible. fout

doc. : IEEE 802. 15 -TG 2_385 r 0 Definitions • Nmin = minimum number of good channels which must be used under FCC • NB = number of bad channels – NBR = number of bad channels to be replaced – NBP = number of bad channels to be grouped/paired • NG = number of good channels Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 Frequency replacement RANDW 12. fgen 7 Mod NG fadp Good channel bank • Used in all circumstances • bypass if NB = 0 (no bad channels) Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 Grouping / Pairing fadp Grouping / pairing fout • Only used if NG < Nmin • bypass block completely otherwise Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 Integration • Proposal TG 2_382 r 0 shows merge is possible – integration is definitely possible; – however, design should keep modularity of systems • Why modular? – allows for implementation of Frequency Replacement only under low power mode (reduced complexity of systems); – allows vendor to choose if to implement both or one of the blocks; – current FCC restrictions on high power require grouping/pairing, future changes may occur; – simplifies block diagram for readability; – splits OPNET simulation work between groups appropriately; Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 Gain scenarios • The following slides show: – Gain of the merged scheme under • Low Power (FCC 15. 249) • High Power (FCC 15. 247 with NPRM) Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 Definitions Transaction #1 Transaction #2 Transaction #3 M S S M f 1 f 2 f 3 f 4 f 5 f 6 • Probability of erroneous transaction P (Bad, Bad) P (Bad, Good) + P (Good, Bad) – Normally: Simple binomial distribution – Grouping/Paring: (see 802. 15 -01/082 r 1) Assumed: Uniform pseudo-random sequence (as in Bluetooth) Interference due only to bad channels 79 Channel mode Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 Additional complexity under FCC 15. 249 or 15. 247 with NPRM • Nmin = 15 (under FCC NPRM) • Example for 23 bad channels (100% duty) – Frequency Replacement (replace 23 channels) • before: P (erroneous transaction) ~ 48% • after: P (erroneous transaction) ~ 0% • decrease error probability by 48% – Grouping/Pairing (remaining 0 channels) • not needed Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 Additional complexity under FCC 15. 249 or 15. 247 with NPRM • Nmin = 15 (under FCC NPRM) • Example for 70 bad channels (100% duty) – Frequency Replacement (replace 64 channels) • before: P (erroneous transaction) ~ 98% • after: P (erroneous transaction) ~ 32% • decrease error probability by 66% – Grouping/Pairing (remaining 6 channels) • before: P (erroneous transaction) ~ 32% • after: P (erroneous transaction) ~ 19% • decrease error probability by 13% – what gain justifies implementation of the scheme? P. T. O. . . Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 Summary of modular merge • Allow vendor to choose if they want to implement both or one of either sub-block; • Both modules have value depending on: – FCC situation – power mode and – vendor’s requirements Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 79 bits for LMP commands • Only transmit ‘good’/’bad’ channel information, not ‘good’/’bad’/’unused’ – – Fits into a normal LMP data unit; doesn’t require special LMP packets to be defined; only requires a new op-code to be defined; More logical. • Solution – Use replacement for first (N-Nmin) bad channels in LMP command – Use grouping/pairing for remaining bad channels Submission

doc. : IEEE 802. 15 -TG 2_385 r 0 LMP Example Assume 70 ‘bad’ channels, 9 ‘good’ channels: HD 1, 1, 1, 0, 1, 0, 0, 1, 1, 1 - - - 1, 1, 0, 0, 1, 1, 0 Replace first (N - Nmin) = 64 channels Assume Nmin = 15 Submission Group/pair remaining 6 channels