Liang Gong

Integrating network function virtualization with SDR and SDN for 4G/5G networks

The rapidly diversified market demands have presented a huge challenge to the conventional mobile broadband network architecture. On one hand, the limited machine room space and insufficient power supply make it impossible to accommodate exponentially growing amount of network equipment of operators. On the other hand, net heterogeneity caused by different specifications of wireless access equipment causes costly trouble related to management and optimization. This article, correspondingly, proposes a holistic solution involving different technologies, i.e. network function virtualization (NFV), software defined radio (SDR), and software defined network (SDN). In particular, we investigate both existing standards and possible extensions for 4G/5G mobile networks, followed by a few open issues for future research.

An intelligent SDN framework for 5G heterogeneous networks

In fifth-generation (5G) mobile networks, a major challenge is to effectively improve system capacity and meet dynamic service demands. One promising technology to solve this problem is heterogeneous networks (HetNets), which involve a large number of densified low power nodes (LPNs). This article proposes a software defined network (SDN) based intelligent model that can efficiently manage the heterogeneous infrastructure and resources. In particular, we first review the latest SDN standards and discuss the possible extensions. We then discuss the advantages of SDN in meeting the dynamic nature of services and requirements in 5G HetNets. Finally, we develop a variety of schemes to improve traffic control, subscriber management, and resource allocation. Performance analysis shows that our proposed system is reliable, scalable, and implementable.

Improve the Performance of LDPC Coded QAM by Selective Bit Mapping in Terrestrial Broadcasting System

In this paper, we employ selective bit mapping to improve the performance of the LDPC coded QAM scheme for terrestrial DTV broadcasting system. The threshold of message-passing decoding can be considerably lowered by selectively mapping the binary components of LDPC codeword to the positions in the m-tuples to be mapped into 2mQAM symbols. In our approach, the mapping pattern is described by bit-mapping polynomials, based on which density evolution can be applied. The optimization algorithm is developed with two implementation concerns, using the Chinese DTMB standard as an example. Numerical results illustrate that our proposed approach can improve the decoding threshold by 0.05 dB to 0.499 dB depending on the code rate and the order of QAM modulation. Simulation results show that the actual BER improvement varies from 0.09 dB to 0.6 dB with different code-modulation combinations in both single-carrier and OFDM modes.

An iterative decoding technique and architecture for RS concatenated TCM coding systems

In this paper, we propose a soft value modification (SVM) algorithm, based on which we further propose a new low-complexity iterative decoding technique for the conventional serial concatenated coding system which consists of Reed-Solomon (RS) code and Trellis coded Modulation (TCM) scheme with an interleaver in between. In our decoding technique, the RS decoder is based on Berlekamp-Massey (BM) algorithm, and the TCM decoder is based on Viterbi (VB) algorithm. After each iteration, the data bits and the err-flag information generated from the output of RS decoder can be fed back to the input of VB decoder as reference information, which combined with the SVM algorithm helps the VB decoder to decide whether and how to modify the corresponding soft values from the channel output. This modification approach makes the soft value more reliable than the previous version and, thus, increases the VB and RS decoders correction capacity. Extensive simulation results in different systems highlight the proposed schemes coding gain over the conventional two-stage decoding algorithm. Finally, a low-complexity architecture for hardware implementation of this novel iterative decoding technique is suggested, and the comparison with the pipeline architecture found in the literature shows the advantages of our proposal.

A Stackelberg Game Based Inter-tier Spectrum Sharing Scheme for LTE-A SON

This paper proposes a spectrum sharing scheme based on Stackelberg game exploiting the SON features of 3GPP LTE-A HetNets. In the game, the MeNB controls and prices the available spectrum resource which the PeNBs can purchase and use to serve the Pico-cell as well as offload some macro-users. Both kinds of nodes try to maximize their own utility which is consisted of user data rates, earning and expenditures on spectrum trading. During the dynamic interaction of the game, the interference coordination features of the LTE-A SON, such as almost blank sub-frames (ABS) of enhanced inter-cell interference coordination (eICIC) and cell range expansion (CRE), are employed so that the change of spectrum allocation can be fully exploited. Simulation results show that the proposed approach can significantly improve the throughput of victim users with slight network total throughput degradation.

Improved shortening algorithm for irregular QC-LDPC codes using known bits

Shortening is a technique to achieve rate and length adaptable low-density-parity-check (LDPC) codes. Other than the puncturing and extending techniques, shortening aims to remove certain information columns, i.e., the columns standing for the known information bits, from a given parity check matrix. In this paper, we first review the recently suggested largest-extrinsic-sum (LES) algorithm which is especially designed for quasic-cyclic (QC) LDPC codes, and then propose a method based on smallest-row-variance priority (SRVP). Moreover, we obtain both theoretical analysis and simulation results of the codes in 802.11n and 802.16e standards, which shows that: 1) the proposed SRVP algorithm always achieves better Eb/No threshold than the existing LES algorithm and generally, the shortened codes outperform the original codes; 2) severe performance degradation occurs when too many information bits are shortened for low rate codes. The proposed algorithm as well as the corresponding performance evaluations is instructive to practical applications.

Multi-rate LDPC codes for DTV transmission by known-bits padding

Modern DTV broadcasting systems, such as DVB and the Chinese DTMB system, usually have multiple code rates for forward error correction to accommodate a variety of channel conditions. In these systems, each code rate corresponds to a special set of code words, which put a limit on rate granularity and thus cannot achieve seamless code rate adaptation. To solve this problem, we propose a known-bits padding scheme to generate multi-code-rate LDPC codes of arbitrary granularity. In our scheme, we take a standardized LDPC code of specific rate as mother code, then pad known-bits according to the targeted code rate and given shortening optimization algorithm. Simulation study shows that our proposed scheme can obtain satisfying performance in terms of code rate flexibility and decoding threshold.

Selective bit mapping for Chinese DTMB LDPC coded QAM modulation scheme

In this paper, we introduce selective-bit mapping to improve the performance of the LDPC coded QAM modulation scheme for Chinese Digital Terrestrial Television Broadcasting (DTMB) system. The threshold of message-passing decoding can be lowered by matching the binary components of LDPC codeword to the positions in the m-tuples to be mapped into 2mQAM symbols. By our approach, the pattern of mapping is described by bit-mapping polynomials, based on which density evolution using Gaussian approximation can be applied. We optimize the bit mapping polynomials for the LDPC codes of the Chinese DTMB standard. Numerical results illustrate that by the proposed approach the decoding threshold can be lowered by 0.05dB to 0.499dB by different combinations of code rate and order of QAM modulations. Simulation results show that the actual BER improvement varies from 0.09dB to 0.6dB in single-carrier and OFDM modes.

A Stackelberg Game Based Inter-tier Spectrum Sharing Scheme Exploiting LTE-A SON Techniques

This paper proposes a spectrum sharing scheme based on Stackelberg game exploiting the SON features of 3GPP LTE-A HetNets. In the game, the MeNB controls and prices the available spectrum resource which the PeNBs can purchase and use to serve the Pico-cell as well as offload some macro-users. Both kinds of nodes try to maximize their own utility which is consisted of user data rates, earning and expenditures on spectrum trading.During the dynamic interaction of the game, the interference coordination features of the LTE-A SON, such as cell range expansion (CRE), is employed so that the change of spectrum allocation can be fully exploited. Simulation results show that the proposed approach can significantly improve the throughput of victim users with slight network total throughput degradation.

Iterative rake structural decision feedback equalization in linear single frequency network of Chinese High Speed Railway Television system

The Chinese High Speed Railway Television (HSRT) has become a new hotpot among applications of broadcasting technology. In this paper, we consider a kind of network infrastructure based on signal frequency network (SFN) design. The character of linear SFN in the railway television system is that receiver between two transmitters will receive signal from both base stations. When single carrier system is adopted as the modulation/demodulation technique, decision feedback equalization (DFE) is used to combat the interference brought by SFN and reflected signals. However, traditional DFE may waste the signal power of dispersive channels. Thus it is not the most effective method in the linear SFN of railway television. To deal with this problem, we propose an iterative rake structural DFE algorithm with cancellation (IR-DFE-C). Theoretical analysis and simulation show that IR-DFE-C algorithm utilizes the power of multipath channels more efficiently than traditional time domain decision feedback equalization (TD-DFE). Moreover, this IR-DFE-C will not be affected by the deep fading problems in frequency domain. Therefore it has advantages over both TD-DFE and frequency domain linear equalization (FD-LE) method which has the problem of noise enhancement in such circumstances.