Most contemporary modems provide a bit-oriented service at the physical layer of the OSI reference model. They do not provide any specific support for packet data transmission. A novel Hop-Aligned Slow Frequency Hopping (HA-SFH) concept is proposed to provide a packet traffic service in a Partial Band Jamming (PBJ) environment. Frequently, SFH is used together with deep interleaving and powerful Forward Error Correction (FEC) to combat PBJ. The proposed HA-SFH exploits the error control mechanism of the link layer packet protocol and reduces the data redundancy by using a low gain FEC, it results in a more bandwidth efficient system. The HA-SFH and deep interleaving SFH (DI-SFH) performance were analysed and compared by using a series of simulations and experiments employing a network simulator and hardware such as Viterbi decoders and a channel noise simulator, etc. The results are presented in terms of throughput and average frame delay, which are more meaningful qualities than BER for packet traffic. It shows concrete results that HA-SFH performs 2 times better than DI-SFH in terms of throughput under PBJ. This thesis also discusses a wide range of technical issues involving the implementation of HA-SFH such as the coding level required, protocol enhancement, networking traffic, adaptive FEC, channel state estimation, etc. It also explores an adaptive code rate system called the Smart Codec (SmCodec), which can be integrated into HA-SFH. The further development of the SmCodec also makes it a suitable system for use in commercial PSK modems to combat the effect of channel fading (eg. from rain). A new channel state estimation technique is also proposed. It estimates the channel state from the block error rate and the required estimation time is as low as 2 seconds for the most required range of E<sub>b</sub>/N<sub>o</sub> at a 64 kbps link (the estimation time decreases when the link speed increases).
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:388941 |
| Date | January 1997 |
| Creators | Pang, Siong Loon |
| Publisher | University of Aberdeen |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://digitool.abdn.ac.uk/R?func=search-advanced-go&find_code1=WSN&request1=AAIU090544 |
Page generated in 0.0017 seconds