In order to improve the quality of video transmission in heterogeneous networks, a method based on lattice quantization for heterogeneous video network joint source channel coding is proposed. First, a multi-descriptive independent parallel channel transmission framework for Gaussian video source transmission is established by using a parallel source transmission scheme for signal sources that describe multiple joint source channel coding; secondly, the use of a lattice-scale quantization method to consider the presence or absence of information decoders, respectively Under the analog mapping bandwidth reduction, through the bandwidth expansion to improve heterogeneous network video transmission performance; Finally, through the three algorithms with VPS, FMS and DMUS encoding redundancy, encoding rate, peak signal to noise ratio, end to end video The comparative experiments on frame delay and effective loss rate indicators verify the effectiveness of the proposed method.
introduction
In recent years, video transmission based on mobile technology (such as online game live broadcast, sports broadcast, etc.) has become a popular streaming media application, and video transmission traffic has also grown rapidly [1]. Video traffic accounts for about 58% of total traffic in 2012 and about 70% in 2017. In 2012~2017, the total mobile traffic will increase by about 14 times[2]. How to ensure the reliability of ultra-clear video data transmission in existing networks is a major challenge faced by service providers.
Although the current network infrastructure construction has provided users with different methods for Internet access, the data transmission capacity of a single network still has limitations and cannot provide satisfactory mobile video transmission performance [3]. The main problem of WLAN networks is limited bandwidth and small coverage, which can not meet the mobile video service requirements of a large number of mobile users. The WiMAX network can provide relatively wider coverage and higher peak rate, but when the amount of users is large, the real-time high-throughput demand for multimedia cannot be met. The limitation of the single network performance makes the heterogeneous network bandwidth integration issue highly valued by scholars. The main research contents of the Joint Source-Channel Coding (JSCC) are the optimization of the rate of channel coding and source, and the error correction coding and channel status of video data: (1) Channel coding and source Coding ratio optimization, for example [5, 6]; (2) How to adjust the coding rate to achieve the required transmission target in the case of setting channel parameters and states, for example, literature [7]; (3) channel coding Reliability improvement, such as Fountain Code, Turbo, etc.; (4) Joint coding optimization algorithm design to achieve system performance improvement, as in [8]; literature [9] studies the optimization of coding rate under the condition of bandwidth limitation. Minimize video end-to-end distortion.
However, the above algorithm uses the error control method, but does not consider the problem of the failure of the channel itself, and it easily leads to the weakening of the transmission quality of the transmission data video, and affects the user experience and the video service quality. In this regard, this article mainly focuses on heterogeneous network video transmission problems. In the decoder, a multi-description joint source channel coding method is studied using lattice quantization conversion. The purpose is to obtain a low-distortion and low-delay video transmission scheme.
1 Problem Model Description
A parallel source channel transmission scheme using a heterogeneous source channel coding (MD-JSCC) is shown in FIG. 1 .
The rate distortion five-tuple (R1, R2, D0, D1, D2) is achievable. Among them, the rate-distortion region R (D0, D1, D2) of the MD-SC problem is a closed-end set of rate pairs (R1, R2). The presence or absence of side information (SI) in this area is unknown in advance. The exceptions are quadratic Gaussian MD-SC cases, ie Gaussian source and quadratic distortion functions:
2 Analogue MD-JSCC scheme based on lattice quantizer
2.1 Multiple Description Simulation Map
2.2 The lattice quantizer
The l-dimensional lattice is a discrete subgroup of Euclidean space and can be described as:
2.3 Coding-Decoding Process
Where θ is the rotation angle. The maximum correlation of the output can be obtained using the rotation angle θ=π/2, see [9]. The proposed joint analog source channel MD encoding scheme is shown in Figure 2.
In Figure 2, (S1, S2) can encode X1 by the following steps:
(2) Case 2: (with SI available encoder) will be coded in case 1, except for the initial pair of auxiliary variables (Z1, Z2)0, the alternative form is:
3 experimental analysis
3.1 Experiment Settings
Exata is selected as the network simulator and is set as follows: The simulation platform version is Exata 2.1. This platform is a high-level simulation version developed by QualNet and can be used for experimental simulation in a semi-physical environment. In order to realize the purpose of obtaining H.264 real-time video stream, use Exata 2.1 and the algorithm of this article to carry on the integrated development. The specific development details can refer to the Exata user manual. In the network structure design, the wired network access port is reserved, and the assigned wireless network interfaces include a WiMAX interface, a WLAN interface, and an HSDPA interface. Through the IP address binding can establish the client and server connections. Heterogeneous network related parameters are shown in Table 1.
This paper compares the proposed SCLQ-JSCC algorithm with the following multi-path/heterogeneous network video transmission solutions: (1) Virtual Path System (VPS). The method uses fountain code to construct the video transmission path of heterogeneous network. In the algorithm implementation, the parameter update period is 0.5 s. The fountain code packet size is 8 B and the symbol length is 512 B. (2) Media Flow Rate Allocation (MFRA). This strategy adopts the maximum usage rate algorithm to optimize the coding rate and redundancy of multi-path video transmission. The number of video layers is set to a value of 1 because of SVC/H.264 coding scalability.
3.2 Analysis of results
In order to verify the performance of the proposed SCLQ-JSCC algorithm, we choose the tolerable transmission loss ratio to compare the forward error correction (FEC) redundancy and video coding rate.
Figure 4 shows the comparison of the redundancy of the three comparison algorithms and the video coding rate. From Fig. 4(a), it can be seen that the proposed SCLQ-JSCC algorithm is significantly better than the VPS and MFRA algorithms in coding redundancy indexes. At the same time, the redundancy of the MFRA strategy needs to be considered due to the redundancy optimization problem. Better than VPS strategy. According to Fig. 4(b), we can see that the proposed algorithm is superior to the two selected comparison strategies in terms of video coding rate. The VPS strategy takes into account the problem of virtual path transmission. Therefore, the coding rate is higher than that of the MFRA algorithm.
In the experimental process, the standard deviation, mean value, and instantaneous value comparison results of the PSNR index (Peak Signal-to-Noise Ratio) of video reception are shown in Table 2. The experimental results in Table 2 show that the algorithm in this paper is always higher than the VPS and MFRA algorithms in City and other four groups of video transmission. This indicates that the distortion in the transmission signal is relatively minimal, and the video recovery quality of the algorithm in this paper Relatively better. On the standard deviation of PSNR indicators, the standard deviation of the algorithm is the smallest, which indicates that the proposed algorithm has better video transmission stability.
The result shown in FIG. 5 is the cumulative distribution of frame delays during video transmission. According to the experimental results in Fig. 5, the frame delay of the proposed SCLQ-JSCC algorithm is significantly lower than that of the VPS and MFRA algorithms. This shows the low latency of the proposed algorithm. Although the VPS algorithm takes into account the issue of virtual path transmission, However, after the video frame is lost, the virtual path needs to be re-established, which will affect the transmission delay of the video frame.
Figure 6 shows the comparison of the effective loss rate index in the [30,80]s time period. It should be noted that the PSNR value in the video transmission process is not only related to the loss rate, but also related to the loss of the video frame. Therefore, the indicator It can reflect the quality change of the video transmission process to some extent.
According to Figure 6, we can see that in the effective loss rate index, the algorithm is smaller than the selected VPS and MFRA algorithm. Because the MFRA adopts the video streaming transmission technology, it leads to a higher loss rate than VPS and the SCLQ-JSCC algorithm. . The above experimental results verify the advantages of the proposed algorithm in video data transmission quality and transmission speed.
4 Conclusion
This dissertation proposes a heterogeneous quantization network-based video joint signal source channel coding method based on lattice quantization, establishes a multi-descriptive parallel parallel channel transmission framework for Gaussian video source transmission, uses lattice-scale quantization to reduce the analog map bandwidth, and improves the bandwidth by extending the bandwidth. Network video transmission performance. In the future, the research will mainly focus on multi-access relay channel or multi-hop network MD-JSCC scheme to simulate network topology.
BIOTEPT General purpose AC motors, also known as induction motors, are electric motors that convert AC power into torque (where torque is the measure of a rotational or twisting force) to drive mechanical loads. They are suitable for powering pumps, blowers, conveyors, and machine tools. Split-phase NEMA frame general purpose AC motors are lower efficiency and have a lower starting torque for a softer startup compared to other AC motors. Capacitor-start NEMA frame general purpose AC motors use a capacitor to provide additional power during startup for higher torque. Capacitor-start/run NEMA frame general purpose AC motors are single-phase motors that use a capacitor to provide additional power during startup for higher torque and a separate capacitor to run with higher operating efficiency. Three phase NEMA frame general purpose AC motors and three phase IEC metric motors operate on efficient three phase power with high torque and don`t require start or run capacitors.
General Purpose Motor,General Purpose Ac Motors,3-Phase Nema General Purpose Ac Motors,Split-Phase Nema Frame General Purpose Ac Motors
Ningbo Biote Mechanical Electrical Co.,Ltd , https://www.biotept.com