Instantaneous packet information based on-demand adaptive retransmission for HARQ

Xiaoying Gan, Zhongren Cao
2010-10-01

SCID:  54.1/zqzxqcxu
Hybrid automatic retransmission request (HARQ) is widely adopted in wireless communication systems to improve the link level and system level throughput. Adaptively selecting retransmission based on the received packet can further improve the system performance by reducing the retransmission overhead. To this end, this paper first introduces the concept of instantaneous packet information (IPI). Then, a new metric is proposed to evaluate the quality of received packets based on IPI. The most important feature of the IPI metric is its in-sensitivity to channel fading. Therefore, it can be used in practice, where channel varies significantly. Furthermore, this paper introduces an IPI-metric based HARQ (IPI-HARQ) featuring an on-demand adaptive retransmission. In IPI-HARQ, a receiver is able to evaluate the quality of a received packet and estimate the desired amount of retransmission when the packet has decoding errors. The retransmission adaption in IPI-HARQ includes two aspects: how many symbols and which part of the packet should be retransmitted. The difference between the current IPI and an target IPI is used to estimate how many bits to be retransmitted. The set of bits that can most efficiently increase IPI is selected to be retransmitted and identified by an indicator vector accompanying the NACK feedback. Clustering is used in IPI-HARQ to limit the feedback overhead. The benefit of IPI-HARQ scheme lies in the fine estimation of required retransmission amount and selection of bits with respect to the quality of current received packet. Simulation results show that the proposed IPI-HARQ scheme clearly outperforms conventional HARQ scheme in terms of throughput.
Publication Details
Publication Date
2010-10-01
Journal
Publisher
ISSN
Access Type
Author Information
Authors
Xiaoying Gan
Zhongren Cao
Explore More Research
Use the citation graph to discover related papers and expand your research horizons.
Click any node to explore
Download PDF
100%