Data Forwarding for RS Group IEEE Presentation Submission Template (Rev. 9) Document Number: IEEE S802.16j-08/100r6 Date Submitted: Source: Kanchei (Ken) Loa, Yi-Hsueh Tsai, Yung-Ting Lee, Hua-Chiang Yin, Voice: Youn-Tai Lee, Chun-Yen Hsu Institute for Information Industry (III) Gamini Senarath, Israfil Bahceci, Hang Zhang, Peiying Zhu, Wen Tong, Derek Yu, Sun Sheng, Mark Naden Nortel Networks Shiann-Tsong Sheu National Central University (NCU) Yih-Guang Jan, Yang-Han Lee Tamkang University (TKU) * Venue: IEEE-SA Sponsor Ballot: P802.16j/D6a Base Contribution: IEEE C802.16j-08/100r6 Purpose: Text proposal for P802.16j/D6a Draft Document. Notice: This document does not represent the agreed views of the IEEE Working Group or any of its subgroups. It represents only the views of the participants listed in the “Source(s)” field above. It is offered as a basis for discussion. It is not binding on the contributor(s), who reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any modifications thereof, in the creation of an IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEE’s sole discretion to permit others to reproduce in whole or in part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this contribution may be made public by IEEE Patent Policy: The contributor is familiar with the IEEE-SA Patent Policy and Procedures: and. Further information is located at and.

Data Forwarding Scheme in an RS Group Defined in 16j/D6a When CID-based forwarding is used for –Non-selective forwarding: all RSs members shall transmit according to the MAPs received, without using the per CID transmission –Selective forwarding: using RS_Member_List_Update message to update CID list at each RS, each RS would look for the data bound to its subordinate stations or data coming from the subordinate stations in the uplink and forward in the assigned times indicated in the MAP.

Issues of Non-selective CID-based Forwarding in DL(1) MAC PDUs are concatenated into a burst by MR-BS and then multicasted to all member RSs. Each RS then partitions the burst into separate PDUs by sequentially decoding each MAC PDU header –RS group members may decode different set of “correct” PDUs when some PDUs are in error due to interference on the relay links –As a result, the RS group members will transmit different burst “content” at the same allocation to an MS  Strong interference at MS

Issues of Non-selective CID-based Forwarding in DL(2) Alternatively, the RS member which detects the PDU errors stops its normal DL forwarding/transmission operations related this burst containing one or more error PDUs. But it causes the following open questions –Does the RS member need to send pilots with null data? If yes, it would degrade the channel estimation performance at the MS. If no, there would be no DL pilot when all RS members detect at least one PDU error

Issues of Non-selective CID-based Forwarding in UL As defined in 16j/D6a, there would be more than one RS member designated to send UL data to the MR-BS in the same allocations Two issues shall be resolved –Similar to DL, RS members may send different PDUs to the MR-BS when errors happen on the access link, which will cause strong interference at the MR-BS –How to perform ranging procedure, especially the UL power control, when different RS member subgroups are designated to transmit data to the MR-BS in the same relay zone?

Issues on Selective CID-based Forwarding In line 53 of page 66 of 16j/D6a, it defines that the RSs receiving the RS_Member_List_Update message may respond with an MR_Generic_ACK or MR Acknowledgement Header for confirmation –ACK only since there is no way for the RS to send NAK

Issues Raised by Using RS_Member_List_Update Message (1) If the BS knows the processing latency required by the RS, it can poll the RS for the RS to send the ACK message or header with minimum delay –T1 >= (PHY + MAC) processing time –Typically, T1 >= 3 frame (1 frame for Rx message +1 frame for MAC processing + 1 frame for Tx ACK message) RS MR-BS RS_Member_List_Update MR_Generic_ACK, or MR_Acknowledgement Header T1 >= 3 frame Frame Number Action

Issues Raised by Using RS_Member_List_Update Message (2) The minimum Frame Number Action = T1 (PHY+MAC processing time) + T2 (Frame Number Action cancellation) T2 >=3 frame (1 frame for BS MAC processing + 1 frame for RS Rx message + 1 frame for RS MAC processing) RS(s) MR-BS RS_Member_List_Update T1 >= 3 frame Frame Number Action X Not Rx MR_Generic_ACK or MR_Acknowledgement Header RS_Member_List_Update T2 >= 3 frame

Issues Raised by Using RS_Member_List_Update Message (3) Member_List_ Update Assign MS to RS1 -actiion time: K1 frames later MR-BS RS1 RS2 MS MS served by RS2 ACK No ACK Message from any RS (data or ack lost ) Member_List_Update Assign MS to RS1 -action time: K2 frames later Frame at which first update message is effective Frame at which retransmitted update message is effective For 5 ms frame duration, 50 ms is 10 frames. Here delay can go frames which is not desirable for delay sensitive services. ACK MS served by RS1 RS2-MS connection gets very weak and RS1-MS connection gets strong

Why Self-contained Burst Transmission is Necessary for RS Grouping (1) When an MS moves from one member RS area to other RS coverage area, the membership update message is sent to the RS group. This message needs to be acked by all the involved RSs in order to take this change into effect. For this, there is an action time (say 5-10 frames). If an ack is not received from all the involved RSs, the membership update needs to be extended to another 5-10 frames.

Why Self-contained Burst Transmission is Necessary for RS Grouping (2) During that time the services to MS in transition is interrupted for their services. Current handover transition interruption is 50 msec in the profile. Ack failure may happen at MR-BS, and message may be lost in the forward link by one of many RSs. Therefore, we should define a forwarding scheme which does not require the membership list so that RSs can forward the data with zero service disruption. The scheme is Burst-based Forwarding Scheme.

Burst-based forwarding scheme In order to solve the data forwarding and signaling overhead problem in relatively faster varying environments as well as the inconsistent data forwarding problem Including a DL_Transmit_Reference IE in the MAP is to provide the selective forwarding information to RSs. The overhead incurred based on this method does not depend directly on the number of connections, but rather depends on number of bursts which can be minimized by proper resource allocation methods.

Summary In RS_Member_List_Update without burst-based forwarding scheme, the action time should be extended to 11 frames in total if the RS_Member_List_Update or ACK is lost at even one of the RSs. (5 frames for detecting the lost and 6 frames for resetting the frame number action time) The burst-based forwarding scheme optimizes the control signaling overhead for resource allocation on the R-link and enables efficient data forwarding from transparent RSs and/or RS group members to their subordinate MSs. RS_Member_List_Update with burst-based forwarding scheme imposes 0 frame latency Adopt the contribution C802.16j-08/148