PHY1- New coding and modulation schemes
Tuesday, 19 June 2018, 11:30-13:00, E1 hall
Session chair: Tomaž Javornik (Jozef Stefan Institute, Slovenia)
11:30 – Turbo-FSK, a Physical Layer for LPWA: Synchronization and Channel Estimation
Francois Dehmas (CEA-Leti Minatec, France); Valérian Mannoni (CEA, France); Vincent Berg (CEA LETI, France)
Turbo Frequency Shift Keying (Turbo-FSK) has been considered as a promising physical layer for low power wide area applications. Its constant envelope at the transmitter combined with performance close to the Shannon’s limit enable to achieve a high energy efficiency. However, results published so far in the literature for this waveform have assumed perfect synchronization and channel estimation. This paper, presents a synchronization and channel estimation approach based on a specifically built preamble and adapted to the performance of the new modulation. Simulations have been performed for both time and frequency synchronization as well as channel estimation. Less than 1 dB degradation in comparison to perfect detection is achieved for the most severe types of channels.
11:48 – A Polar Code Hybrid Rate Matching Scheme
Fengjun Xi, Chanxuan Ye and Robert L. Olesen (InterDigital Communications, Inc)
In recent 3GPP NR standards discussions for 5G, polar codes have been adopted as the control channel coding scheme when the payload size is more than 11 bits. The rate matching schemes for polar codes include shortening, puncturing and repetition. Each of these rate matching schemes has a performance advantage for different conditions. Hence, it is beneficial to include all three types of rate matching schemes. A circular buffer is also designed to support these rate matching schemes. In this paper, we propose a hybrid rate matching scheme which incorporates both shortening and puncturing techniques. This hybrid rate matching scheme provides a performance advantage for some conditions. Also, this hybrid rate matching scheme does not cause a significant impact on the existing polar coding system architecture. Simulation results are provided to show the performance benefit of this hybrid rate matching scheme.
12:06 – Achievable Rates for HDF WPNC Strategy with Hierarchical Bit-Wise Network Coding Maps for Higher-Order Constellations in H-MAC Channel with Relative Fading
Jan Sykora and Petr Hron (Czech Technical University in Prague, Czech Republic)
The paper addresses Wireless Physical Layer Network Coding (WPNC) with Hierarchical Decode and Forward (HDF) strategy. We analyze achievable hierarchical rates in one stage hierarchical MAC (H-MAC) channel for higher order component constellations with bit-wise Hierarchical Network Code (HNC) maps. This is motivated by a possible application of state-of-the-art binary (e.g. LDPC) codes over higher order constellations. We show that the bit-mapped binary codes do not have the same achievable rates as direct higher-order codes. On top of this, the individual bits in the HNC map might provide very uneven performance and it strongly depends on the combination of component alphabets. The results are supported by a validation with practical LDPC codes.
12:24 – Quantifying the Gain of Multi-Connectivity in Wireless LAN
Nick Schwarzenberg, Albrecht Wolf, Norman Franchi and Gerhard Fettweis (Technische Universität Dresden, Germany)
Multi-Connectivity (MC) using multi-link diversity is a promising approach to enable Ultra-Reliable Low-Latency Communications. While recent analytical results about MC already indicate improved reliability compared to a single link, we are interested in the actual gain of implementing MC into an established wireless standard. Therefore, we evaluate inter-frequency MC over independent fading channels for the physical layer of Wireless LAN (WLAN). We compare three combining schemes, namely Selection Combining (SC), Maximal Ratio Combining (MRC), and the recently emerging Joint Decoding (JD), by their respective packet error rates in Monte-Carlo simulations. Our results show that MC outperforms a single link for all of the three combining schemes, even if the total transmit power is not increased but shared among the available links. JD performs best and is found to be considerably less dependent on SNR estimation than MRC and SC. We show that packet synchronization in WLAN causes an error floor in frequency-selective channels which can be effectively mitigated by MC.
12:42 – A Distribution-Free Analysis of Outage Probability over Block Fading Channels
Jinho Choi(Gwangju Institute of Science and Technology (GIST), Korea)
In this paper, we derive a non-asymptotic upper- bound on the outage probability for a multiple transmit diversity technique (with a low diversity gain), which is distribution-free. The derived bound can be used to determine key parameters for the initial stage of reliable connection (with a certain target outage probability) when a system operates under an unknown channel environment.