Congestion Control for Vehicular Safety:

Presentation on theme: "Congestion Control for Vehicular Safety:"— Presentation transcript:

1 Congestion Control for Vehicular Safety:
Synchronous and Asynchronous MAC Algorithms Published in: Proceeding VANET '12 Proceedings of the ninth ACM international workshop on Vehicular inter-networking, systems, and applications Authors 이 논문은 김효곤 교수님의 논문에 레퍼런스중 하나로 모티브가 된 논문이라고 생각됩니다. 매우 흡사한 부분이 많았습니다. Sundar Subramanian Qualcomm R&D NJRC, Bridgewater, NJ, USA Marc Werner Qualcomm CDMA Technologies, GmbH, Nuremberg, Germany Shihuan Liu Iowa State University, Ames, USA Jubin Jose Qualcomm R&D NJRC, Bridgewater, USA Radu Lupoaie Qualcomm CDMA Technologies, GmbH, Frankfurt, Germany Xinzhou Wu

2 CONTENT 1. Introduction 2. Performance of IEEE 802.11p MAC
3. DCC Algorithm evaluation and improvement 4. Standard compliant synchronous mac 5. Conclusion

3 Introduction In August 2008, the European Telecommunications Standards Institute (ETSI) allocated 30 MHz of spectrum in the 5.9 GHz band for ITS In Europe, the ETSI organization has standardized a framework to mitigate the MAC layer congestion issues at high vehicle densities, usually referred to as the Distributed Congestion Control (DCC) algorithm 하지만 문제점이 많다는 것입니다.

4 Performance of IEEE 802.11p MAC
The IEEE p functionalities were simulated by using the 80211MacExt package in ns2 우선 이논문에서는 기본적으로 DCC가 쓰이지 않았을때 p 성능분석을 위해 시뮬레이션을 하였습니다. 6개의 레인이 4미터씩 총 2000미터.. 차량의 개수를 변화시켜 성능을 알아보았다.

5 Performance of IEEE 802.11p MAC
차량밀도가 올라갈수록 Aloha와 매우 비슷해짐. CW가 올라가면 디바이스가 같은 backoff counter를 고를 확률을 줄어들지만 결국 기다리는 시간도 증가하는 것이므로 성능에는 큰영향이 없었다. Does not guarantee any ‘guard zones’ around a transmitter (region where no other transmitters are allowed to transmit simultaneously). Observe that the limitations of IEEE p result in the packet reception performance tending towards an undesirable ALOHA-like behavior.

6 DCC Algorithm evaluation and improvement
ETSI algorithm for decentralized congestion control The transmission parameters associated with a state may include transmit power (P), packet transmission interval (PI) and carrier sense threshold (CST), coding scheme (MCS) among other parameters. The channel load (CL) is defined to be the fraction of time the received power was greater than the CST. 그에 반해 ETSI가 제안 DCC는 State를 세개로 나누어서 congestion을 컨트롤했다. 혼잡해질수록 datarate를 올린다.

7 DCC Algorithm evaluation and improvement
Performance evaluation of the ETSI DCC algorithm 거의 모든 노드들이 거의 ACTIVE에 몰려있음.. 300m까지만 보장.

8 DCC Algorithm evaluation and improvement
Performance evaluation of the ETSI DCC algorithm

9 DCC Algorithm evaluation and improvement
Improving vehicular discovery under DCC 표준에서도 Active 상태를 여러 개로 할수 있다고 했습니다.

10 DCC Algorithm evaluation and improvement
덴서티가 600 에서는 잘 동작 Relaxed 상태에 머문 경우가 많음. DCC보다는 나은 성능을 보여주지만 그래도 부족함.

11 Standard compliant synchronous MAC
Obtaining synchronous transmissions using IEEE p MAC Controlling the channel access time Strict synchronization through GPS Since the same GPS receiver can provide synchronization of sub-microsecond level accuracy, it is sufficient for the successful operation of our algorithm. An additional TDM SYNC layer above the existing MAC layer to create fixed resources and by a distributed resource selection (and reselection) algorithm to allocate the TDM resources.

12 Standard compliant synchronous MAC
Simulation Setup and Results

13 Conclusion 802.11p 및 DCC의 문제점을 정확히 시뮬레이션을 통해서 분석하였으나 정성적인 표현에 그침.
제안한 알고리즘은 MAC 의 동작방식을 많이 수정하였고, 설명 부분이 미흡 GPS를 통해 SYNC를 맞추지만 추상적이고 제대로 적당할지 의문 시뮬레이션에서 다양한 토폴로지 및 토폴로지의 변화가 고려하지 않음