QoS and rate enhancement in DVB-S2 using fuzzy rule based system

Abstract

Adaptive Coding and Modulation (ACM) is one of the most promising features in many communication systems like WIFI, WiMAX and DVB-S2, in terms of throughput and link utilization. In this way the transmission parameters like power, code rate and modulation symbol are adapted according to the varying link conditions. In this paper we propose a Fuzzy Rule Based System (FRBS) for an efficient ACM in DVB-S2 standard. We have taken it as a constrained optimization problem in which the link reliability and throughput is maximized, while power constraint is satisfied. To do so we have selected the standard transmission Modulations and Codes (MODCODs) optimally using FRBS, according to the DVB-S2 downlink channel state information (CSI) at different sites. The FRBS module can easily be embedded in the ACM management module (AMM) which is a software module. Significance of the proposed scheme over other schemes is shown through simulations.

Keywords

DVB-S2 FRBS ACM MODCODs DVB-S

1 Introduction

The technique of broadcasting the audiovisual contents via satellite has been used for many decades. This has always been a suitable choice for broadcasting because of high capacity, wide coverage and entertaining million of users simultaneously over the globe.

However, this choice costs much to satellite service providers (SSPs) and ultimately to subscribers, due to high leasing cost of satellite transponders. So the techniques for optimum spectrum utilization are welcomed to broadcast more audiovisual contents. In this regard adaptive communication and audiovisual compression algorithms are playing a vital role. Digital Video Broadcast –Second Generation (DVB-S2) [1] has a 30% increase in efficiency compared to the digital video broadcast first generation (DVB-S) and has built-in support for ACM. However, for further improvement in efficiency, the ACM feature of DVB-S2 yet to be exploited for broadcast services. That is to intelligently adapt the modulation scheme and code rate, namely MODCODs for a specific broadcast service according to link quality feedback from the earth stations.

In [2], the adaptive coding and modulation feature of DVB-S2 was utilized for broadcasting services of different audio video standards like H.264 Advance Video Coding (AVC) for transmission of high definition (HD) contents. Technical viability of the proposed scheme was shown by taking two different tropic zones that is Greece zone and Indonesia Zone. Finally, it was shown theoretically that the increase in efficiency using DVB-S2-ACM is 67% more than DVB-S and 29% more than DVB-S2 with Constant Coding Modulation (CCM). In [3], DVB-S2 adaptive coding and modulation for high altitude platforms (HAPs) is investigated. In this paper, two different operating scenarios were focused for static and mobile channel. The spectral efficiency was achieved by using a subset of standard MODCOD set of DVB-S2 rather than using the whole set.

According to [4], in DVB-S2 specifications it was focused on three concepts that is best transmission performance, total flexibility and reasonable complexity. To achieve these features, best modulation and coding standards were chosen. Low Density Parity Check (LDPC) codes were chosen among seven candidates for forward error correction (FEC) codes. This causes a 30% increase in performance compared to DVB-S.

In [5], techno-economic perspectives of the use of DVB-S2 and its unique feature, Adaptive Coding and Modulation (ACM) in the provision of satellite triple play were studied. It was shown that how ACM can help increase in spectral efficiency for satellite triple play. For this one user per terminal and group of users per terminal were studied. For group of users per terminal WiMAX was considered for redistribution of audiovisual services. In a technical report [6], the viability of DVB-S2 in terms of more users support, efficiency and cost effectiveness were studied for SSPs to lower their operating cost and increase their business opportunities. In the technical report [7], rain fade management in third generation Ka-band satellite systems were presented. It was shown that rain fade can be compensated by suitably using adaptive coding and modulation. Moreover, that ACM can increase network availability even in rain with thunder and storm.

For single antenna OFDM systems, coded bit and power loading problem was addressed by Li et al. using LDPC codes [8] originally motivated by [9]. Many bit interleaved coded modulation (BICM) systems have been proposed like [9 –11]. Lei et al. investigated adaptive communication using turbo codes [12]. An adaptive coding and modulation scheme is proposed by Bockelmann et al. in which a bisection method was used to adapt the transmit rate [13].

Adaptive coding and modulation scheme with fixed transmit power by Atta-ur-Raman et al. [14] proposes a Fuzzy Rules Base System for finding suitable modulation code pairs (MCP) for the subcarriers given the Quality of Service (QoS) demand and channel state information of all subcarriers. Same strategy while using Product Codes as coding scheme was proposed in [15]. A combined approach to adaptive coding, modulation and power was proposed by the authors [16] in which Fuzzy Rule Base System was used for choosing modulation code pair while power vector for subcarriers was optimized by using two different algorithms, Water filling algorithm and GA. Similarly, in [17], same authors proposed differential evolution (DE) in conjuction with FRBS to adapt the transmission parameters in Orthogonal Frequency Division Multiplexing (OFDM) environment. A real time adaptive coding and modulation scheme for OFDM systems was proposed in [18], in which a Gaussian Radial Basis Function Neural Network (GRBF-NN) was trained using FRBS and DE. In these publications channel information was assumed to be known on both sender and reciever ends. However there are approaches where channel estimation, classification based on feature extraction in noisy environment is also a hottest area of research [19, 20].

In this paper, we propose a Fuzzy Rule Based System for efficient adaptive coding and modulation scheme depending upon the reception quality feedback from different sites provided by DVB-S2/RCS and QoS demanded by the type of TV service. In this scheme FRBS is responsible for selection of optimum MODCODs per TV service such that it maximizes the satellite downlink data rate while satisfying the power and target bit error rate constraint.

The remainder of this paper is organized as follows. In Section 2, system model is introduced. Performance of different codes in conjunction with different modulations is presented in Section 3. The results of Section 3 are used in Section 4 to formulate a constrained optimization problem. In Section 5 a brief introduction to Fuzzy Rule Base is given that is used to solve the optimization problem formulated in previous Section. Section 6 presents the performance comparison of proposed scheme with various other famous adaptive schemes while Section 7 concludes the paper.

2 System model

The system model considered for this experiment is conformed to DVB-S2 standard and shown in Fig. 1. Reception Quality Feedbacks (RQF) are sent to ACM Management and Multiplexing (AMM) block via DVB-S2/RCS (Return Channel via Satellite) [2]. Also this return channel might be a terrestrial return channel. These feedback reports are formulated by ACM Reporting Module (ARM) embedded in the probe receivers at different sites. ARM module applies a Simple Network Management Protocol (SNMP) query to respective receiver and derives respective carrier to interference plus noise ratio (CINR) value.

CINR is considered as most appropriate indicator of satellite reception signal quality though it is mostly subject to Additive White Gaussian Noise (AWGN) compared to any other channel impairment. These reports are sent to respective block after a predefined interval T. These reports consist of an average of P number of CINR values obtained during interval T. This can be written as; $< CINR > = {\frac{1}{P} \sum_{i = 1}^{P} {CINR}_{i} |}_{T}$ (1)

AMM is a software module that decides in real time that which MODCOD should be used for which TV service. Proposed Fuzzy Rule Based System is part of this software module. So we name this module as FRBS based ACM management and multiplexing (FRBS-AMM) block. This block generates ACM commands with respect to the feedbacks from different sites for different TV services. These commands are fed to DVB-S2 modulator, which as result modulates and encodes the data of different services to respective MODCOD suggested by FRBS-AMM. We assume that there are N number of services and M number of MODCODs available in the system. These modulated services’ data is sent to different earth stations via DVB-S2 downlink. This sequence can be seen in Fig. 1.

3 Coding and modulation

Components of the proposed system model are described below.

3.1 Coding schemes

Powerful forward error correction codes are used for ACM that are concatenation of BCH (Bose-Chaudhuri-Hocquenghem) codes [24, 25] and LDPC (Low Density Parity Check) codes [26] that result in quasi error free (QEF) performance just 0.7 dB away from Shannon limit and a packet error rate (PER) of 10^-7 depending upon transmission mode. The supported code rates fall in the set C.

$C = {1 / 4, 1 / 3, 2 / 5, 1 / 2, 3 / 5, 2 / 3, 3 / 4, 4 / 5, 5 / 6, 8 / 9, 9 / 10} .$

For decoding, standard message passing algorithm is used for LDPC codes.

3.2 Modulation schemes

As modulation scheme for adaptive modulation are used from the set of standard modulations used in DVB-S2 standard that are optimized for operation over non-linear transponders. These supported modulations are listed in the set M. $M = {QPSK, 8 - PSK, 16 - APSK, 32 - APSK} .$

For experimentation the sequence of operations is carried out in same way as given in Fig. 1. Same components are chosen and the arrangement for performance analysis is shown in Fig. 2. Transmitted signal is firstly encoded by encoding block which consists of BCH codes followed by LDPC codes encoder by using the code rate from the set C. The signal is then passed through the inter-leaver which is followed by the modulator. The interleaved signal is modulated using the elements from the set M. In this way we have following pairs of coding and modulation by cross product of sets C and M, which yields (Equation 2),

$\begin{matrix} MODCOD = C x M \\ = {(c_{i}, m_{j}) | \forall c_{i} \in C, \forall m_{j} \in M} \end{matrix}$ (2)

After passing through Additive White Gaussian Noise (AWGN) channel, a soft demodulator is employed whose output is fed into de-interleaving block. After passing through decoding block consisting of LDPC decoder using message passing algorithm, followed by BCH optimal decoder, bits are received and bit error rate performance is measured.

Then bit error rate performance of each pair is obtained over an Additive White Gaussian Noise (AWGN) channel. The selection of this channel is suitable in a sense that it reflects the proper relationship between signal to noise ratio (SNR) and data rate achievable under a specific target bit error rate (BER). Also other channel characteristics like fading types etc can be compensated easily. Moreover, received signal to noise ratio (RSNR) is considered as a good indicator for selection of MODCOD in DVB-S2, which is directly linked with the impairment offered by the AWGN channel.

The theoretical spectral efficiency of each modulation code pair (MODCOD) can be found in Table 1, which is a product of code rate and modulation bits per symbol. In these MODCODs [QPSK, 1/4] has the minimum spectral efficiency that is 0.5bits/s/Hz while [32APSK, 9/10] has the maximum spectral efficiency that is 4.5bits/s/Hz and there are forty-four MODCODs in total.

It can be seen from the table that there are more than one MODCODs having same spectral efficiencies like both [8PSK, 1/3] and [QPSK, 1/2] have efficiency 1bits/s/Hz. According to [1] the recommended set of MODCODs being used in DVB-S2 standard are twenty-eight. These MODCODs are highlighted in Table 1.

The entries of the above given table can also be viewed in terms of a diagram that shows the relationship between MODCOD being used and the spectral efficiency it exhibits. The shaded entries in the Table 1 represent the practical MODCODs being used in the DVB-S2 standard.

Figure 3 shows the spectral efficiencies of MODCODs in three ways, that is, theoretical, normal frame with pilot signal and normal frame without pilot signal payload. Without pilot payload is the most considerable case since pilot frames are used only as preamble frames.

The bit error rate (BER) performances of all MODCOD combination are plotted over AWGN channel using the sequence shown in Fig. 2. These performance results are shown in Fig. 4.

4 Rate optimization

In order to maximize the data rate for DVB-S2 system, following constrained optimization problem will be considered. $\begin{matrix} max R_{Total} = \frac{1}{N} \sum_{i = 1}^{N} r_{i} \\ s . t, \\ {BER}_{i} \leq {BER}_{{QoS}_{i}} \\ and \\ P_{Total} = \sum_{i = 1}^{N} p_{i} < P_{T} \end{matrix}$ (3)where r_i = (log ₂ (M)) _iR_i is bit rate for the ith service, which is a product of code rate R_i and modulation order (log ₂ (M)) _i used for ith service (means the MODCOD), P_T is the available transmit power and BER_{QoS
_i} is target BER that depends upon a specific quality of service (QoS) demand by the ith service, while N is number of TV services in DVB-S2.

From the results obtained in Section-III, those code-modulation pairs that fulfill different bit error rate demands from different TV services or satellite applications i.e. BER_T = 10^-7, 10^-6, 10^-5, 10^-4, 10^-3 etc are obtained by drawing straight lines on the graphs shown in Fig. 4, against different BER points.

This process is shown in Fig. 5, where there are three circles; just to focus the points of intersection. First circle shows that for obtaining a BER of 10e-2 with a SNR value -3 dB, [QPSK, 1/4] MODCOD can be used. Similarly, second circle shows that for obtaining a BER of 10e-2 and a given SNR value as 2 dB, [QPSK, 3/5] can be used and according to third circle for obtained BER 10e-4 at 3 dB SNR, [QPSK, 3/4] can beused.

As there are twenty-eight curves according to twenty-eight MODCODs and we have drawn lines against the six BER points that is 10^-7, 10^-6, 10^-5, 10^-4, 10^-3, 10^-2 so in this way we obtained one-hundred and sixty-eight (28×6 = 168) points.

So the information obtained can be expressed as “for a given received SNR value and specific QoS demand which MODCOD can be used”. All the points are listed in a table. This table after completion is used as a starting point for generation of look-up table for the fuzzy rule base system.

The steps involved in creation of FRBS are described below with a brief description of each step.

Data Acquisition

Data is obtained from the graphs obtained in Section III in terms of input/output (IO) pairs. This is taken by putting the horizontal lines against various bit error rates and points of intersection with the curves are noted. The data obtained is in the form of “A suitable MODCOD according to a given receive SNR and required quality of service”.

Rule Formulation

Rules for each data pair are obtained by using the appropriate type and number of fuzzy sets. This is done substituting the complete data pair in the fuzzy input output space. Each data pair generates one rule. The procedure in detail can be seen in [23].

Elimination of Conflicting Rule

The rules having same IF part but different THEN parts are known as conflicting rules. This appears when there exist more than one MODCODs against a given set of specification.

Completion of Lookup Table

Look up table is complete in a sense that there exists a rule for every situation.

5 Proposed fuzzy rule based system

A fuzzy rule based system (FRBS) is used to optimize the cost function given in Equation 3. It will be decided that which MODCOD is suitable for a specific TV service based upon the average channel to interference and noise ratio (CINR) at the receiver and the Quality of Service demand. The rules are of theform; $(x_{1}^{j}, x_{2}^{j}; y^{j}); j = 1, 2, 3 \dots \dots N$ (4)where $x_{1}^{j}$ represents received SNR, $x_{2}^{j}$ represents required BER (QoS) and y^j represents the output MODCOD suggested by FRBS, so a rule can be expressed as;

{IF (x₁ is L1 and x₂ is Q7) THEN y is P2}

A brief description of each component of fuzzy rule based system being designed, is presented subsequently. Design of the FRBS is carried out in MATLAB 7.8.0 using standard Fuzzy System Toolbox.

Fuzzy Sets

Sufficient numbers of fuzzy sets are used to cover the input output spaces. There are two input variables received CINR and QoS. There is one output variable that is modulation code pair (MODCOD).

Twenty-one fuzzy membership functions for input variable CINR, six fuzzy membership functions for input variable QoS and twenty-eight fuzzy membership functions for output variable MODCOD are used respectively.

Fuzzifier

Standard Gaussian fuzzifier is used with AND as MIN and OR as MAX. This is shown in Fig. 6–8 for two input variables CINR, QoS and one output variable MODCOD respectively.

Rule Base

Rule base contains rules against all the IO pairs. As there are twenty-one sets (L0 to L20) for first input variable named CINR and six sets (Q2 to Q7) for input variable QoS, hence there are one-hundred and twenty-one rules in rule base. Rule base is complete in the sense that there exists rule for all possible combinations of input spaces.

Inference Engine

Famous Mamdani Inference Engine (MIE) is used that will infer which input pair will be mapped onto which output point.

De-Fuzzifier

Standard Center Average Defuzzifier (CAD) is used for defuzzification due is its accuracy and low computational complexity.

Once the knowledge is pumped into the rule base, fuzzy inference engine (FIE) decides the optimal MODCOD for a given channel state information at a certain TV service site and required quality of service.

The mechanism for throughput calculation is given in Fig. 9. This figure shows that an equal power is transmitted towards all satellite sites running different TV services. Upon receiving the attenuated power and quality of service requests from different services the FRBS will suggest accordingly MODCODs for TV services such that the throughput should be maximized and power constraint can be met. The average throughput calculation using FRBS is given in Equation 5.

$\begin{matrix} R_{Total} = FRBS (QoS, P, α) \\ = MODCODs = \frac{1}{N} \sum_{i = 1}^{N} r_{i} \end{matrix}$ (5)

6 Results

In this section performance of the proposed scheme is demonstrated and compared with various other schemes of the same area, through the simulations and the results are shown in the form figures. The parameters used for the simulations are given in Table 2.

In Fig. 10, performance of proposed scheme is shown for different target packet error rates (PER) like 10e-2, 10e-3, … , 10e-7 for the different supported TV services. In this figure the graphs show the spectral efficiencies versus received signal to noise ratio (SNR) for fixed PER. It is apparent that if the PER is relaxed that is say 10e-2, the spectral efficiency approaches near 4.5bits/s/Hz and for high quality (say PER = 10e-7) achievable efficiency approaches to 1.8bits/s/Hz. Similarly, the high values of SNR result in better spectral efficiencies and vice versa.

In Fig. 11, upper and lower bound on the performance of the proposed scheme is shown. In this figure upper bound is given to scenario when QoS is low that is with the 10e-2 PER and in this case the spectral efficiency or throughput approaches to its maximum. Similarly, the lower bound is given to the scenario when QoS is very stringent that is with PER as 10e-7. In this case, the spectral efficiency approaches to its minimum value. Third graph is obtained by taking an average case into consideration in which each TV service is demanding a different QoS at random. For this case the throghput approaches to its average value that is 3.25 bits/s/Hz.

In Fig. 12, the proposed scheme is compared with another scheme presented by [3]. In the scheme proposed by [3], the MODCOD adaptive selection criteria was based on few predefined fixed thresholds of received SNR. That is if recieved SNR will fall in a range a specific MODCOD will be used and so on. The first limitation of that scheme was insufficient number of MODCOD support due to few SNR thresholds. Secondly, the scheme did not take into account the QoS demand from different TV channels. Result shows that proposed scheme performs significantly better than the previous scheme even in low SNR regions. This is because in [3], authors have utilized MODCOD subsets instead of examining the whole set of supported MODCODs, while in our proposal FRBS can choose any supported MODCOD depending upon the situation.

In Fig. 13, the proposed scheme is investigated in case of video service broadcast where required spectral efficiency is between 2bits/s/Hz to 2.8bits/s/Hz. In [2] fixed operating thresholds and selected MODCODs are adopted without any consideration of QoS demand from the user. Proposed scheme performs significantly better than [2], where optimum MODCODs are chosen according to the desired QoS and received SNR from different earth stations.

7 Conclusions

In this paper a fuzzy rule based system is proposed for adaptive coding and modulation (ACM) in a digital video broadcast second generation (DVB-S2). As DVB-S2 supports ACM, our proposal can be incorporated as a software module in ACM management block in DVB-S2. Proposed scheme takes quality of service demand and received SNR from different TV services into account to choose optimum MODCOD for the specific TV service. It is shown through simulation that proposed scheme performs better than well-known schemes in the literature. The performance of proposed scheme is implemented in MATLAB 7.8.0.

Footnotes

Acknowledgments

This research work was supported by Higher Education Commission (HEC), of Pakistan.

References

ETSI EN 302 307, Digital Video Broadcasting (DVB); Second generation framing structure, channel coding and modulation systems for Broadcasting, Interactive Services, News Gathering and other broadband satellite applications. ETSI European Standard, v.1.1., 2005.

Gardikis , Zotos

and Kourtis

, Satellate media broadcasting with adaptive coding and modulation, International Journal of Digital Multimedia Broadcasting, Hindawi Press, vol. 2009, 2009.

Smolnikar

, Javornik

, Mohorcic

and Berioli

, DVB-S2 Adaptive Coding and Modulation for HAP Communication, IEEE Vehicular Technology Conference, 2008, pp. 2947–2951.

Morello

and Mignone

, DVB-S2: The second generation standard for satellite broad-band services, Proceedings of the IEEE 94(1) (2006), 210–227.

Anastasiadou

, Gardikis

, Nikiforiadis

and Pangalos

, Adaptive Coding and Modulation-enabled satellite triple play over DVB-S2 (Digital Video Broadcasting - Satellite - Second Generation): A techno-economic study, International Journal of Satellite Communication and Neworking 30(3) (2012), 99–112.

White Paper, iDirect Evolution DVB-S2/ACM System, 2008. www.idirect.net

Techniacl Report, Rain Fade Management in 3rd Generation Ka-band Satellite Systems, KaComm Communications Pty Ltd, 2009.

and Ryan

W.E.

, Mutual-information-based adaptive bit-loading algorithms for LDPC-Coded OFDM, IEEE Trans on Wireless Communications 6(5) (2007), 1670–1680.

Caire , Taricco

and Biglieri

, Bit-interleaved coded modulation, IEEE Trans on Information Theory 44(3) (1998), 927–946.

10.

Stierstorfer

and Fischer

R.F.H.

, (Gray) Mapping for Bit-Interleaved Coded Modulation, in IEEE Vehicular Technology Coference (VTC2007-Spring), Dublin, Ireland, 2007.

11.

Stiglmayr

, Bossert

and Costa

, Adaptive Coding and Modulation in OFDM Systems using BICM and Rate-Compaitble Punctured Codes, in European Wireless, Paris, France, 2007.

12.

Lei

and Burr

, Adaptive Modulation and Code Rate for Turbo Coded OFDM Transmissions. Vehicular Technology Conference VTCpp, 2007, pp. 2702–2706.

13.

Bockelmann

, Wübben

and Kammeyer

K.D.

, Rate Enhancement of BICM-OFDM with Adaptive Coded Modulation via a Bisection approach, IEEE 10th Workshop on Signal Processing Advances in Wireless Communications,SPAWC ’09, 2009, pp. 658–662.

14.

Atta-ur-Rahman , Qureshi

and Malik

, A fuzzy rule base assisted adaptive coding and modulation scheme for OFDM systems, J Basic Appl Sci Res 2(5) (2012), 4843–4853, 2012. [ISI Indexed Journal].

15.

Atta-ur-Rahman , Qureshi

and Muzaffar

, Adaptive coding and modulation for OFDM systems using product codes and fuzzy rule base system, International Journal of Computer Applications (IJCA) 35(4) (2011), 41–48.

16.

Atta-ur-Rahman , Qureshi

and Malik

, Adaptive resource allocation in OFDM systems using GA and fuzzy rule base system, World Applied Sciences Journal 18(6) (2012), 836–844. [ISI Indexed Journal].

17.

Atta-ur-Rahman , Qureshi

, Malik

and Naseem

, Dynamic resource allocation for OFDM systems using differential evolution and fuzzy rule base system, Journal of Intelligent & Fuzzy Systems (JIFS) 26(4) (2014), 2035–2046.

18.

Atta-ur-Rahman , Qureshi

, Malik

and Naseem

, A real time adaptive resource allocation scheme for OFDM systems using GRBF-neural networks and fuzzy rule base system, International Arab Journal of Information Technology (IAJIT) 11(6) (2014), 593–601.

19.

Gai

Shan

, Yang

Guowei

and Zhang

Sheng

, Multiscale texture classification using reduced quaternion wavelet transform, International Journal of Electronics and Communications 67(3) (2013), 233–241.

20.

Gai

Shan

, Yang

Guowei

, Wan

Minghua

and Wang

Lei

, Denoising color images by reduced quaternion matrix singular value decomposition, Multidimensional Systems and Signal Processing 26(1) (2013), 307–320.

21.

ETSI EN 301 790, Digital Video Broadcasting (DVB); Interaction Channel for Satellite Distribution Systems, ETSI European Standard, 2000, v. 1.3.1, 2003.

22.

CCSDS 131.3-B-1, CCSDS Space Link Protocols over ETSI DVB-S2 Standard, The consultativeCommittee for SpaceData Systems, 2013.

23.

Wang

L.X.

, A Course in Fuzzy Systems and Controls. (Chapter 12). Prentice Hall Publications, 1997.

24.

Hocquenghem

, Codes correcteurs d’erreurs, Chiffres (in French) (Paris) 2 1959, pp. 147–156.

25.

Bose

R.C.

and Ray-Chaudhuri

D.K.

, On a class of error correcting binary group codes, Information and Control 3(1) (1960), 68–79.

26.

Robert

G.G.

, Low Density Parity Check Codes. Monograph, M.I.T. Press. 1963, Retrieved August 7, 2013.