Tutorials to .com

Tutorials to .com » Mechine » Display » NS-based network video QoS analysis

NS-based network video QoS analysis

Print View , by: iSee ,Total views: 20 ,Word Count: 2310 ,Date: Wed, 10 Jun 2009 Time: 12:28 AM

With the networks and the rapid development of multimedia technology, network in an endless stream of multimedia applications, such as video conferencing, video on demand (VOD), multimedia services, such as distance education. Image as a result of the activities of a large amount of data in specific applications, network topology, network bandwidth, routing technologies, as well as the actual packet transmission programs and other factors will be divided on the impact of network performance, and ultimately affect the video quality of service. It is essential that the network's QoS video transmission security research.

QoS in the network security issues in the course of the study, because of network complexity and economic conditions, network simulation technology plays a very important role. Simulation technology is a network through the establishment of network equipment and network links the statistical model, and simulate network traffic transmission, in order to obtain the network design or optimization of network performance required for the simulation data. Based on the analysis tools EvalVid based on will be based on the NS-2 network simulation software, a network video QoS analysis.

1 video overview of the QoS research techniques

In the simulation environment, researchers usually use a different yardstick to measure its transmission strategy proposed by the impact of video quality, the scale including the packet loss rate, packet latency, frame loss rate and frame rate, such as decoding. Many of these scales to measure network performance and can not fully reflect the quality of video. In video quality evaluation methods are mainly subjective and objective quality evaluation of two kinds of methods of quality assessment. Subjective quality evaluation is a relatively accurate method of evaluation of image quality, as a direct reflection of the subjective quality evaluation of the feeling of the human eye. Commonly used indicators are based on the quality of 5 or 5 injury average observations system (MOS points). Quality video images through the objective evaluation of certain mathematical criteria, a common peak with peak signal to noise ratio (PSNR). It is defined as follows:

NS-based network video QoS analysis

One, M and N is the video frame width and height dimension; x (i, j) and x '(i, j) is the original and reconstructed pixels in the pixel (i, j) points, the brightness or color values . Although the network QoS research techniques video one after another, but researchers had not been able to give a unified architecture for video quality from the subjective and objective evaluation. Jirka Klaue for this is given in the literature evaluation of video transmission quality of a unified framework and tool set - EvalVid. EvalVid with modular structure, you can use in the video coding strategy used, the network of interaction between the trace file through the achievement, it is easy to experiment in any environment. Chih-Heng Ke, etc. EvalVid in the literature and NS2 will combine to give a qualitative assessment of video streams of the new tool set Enhanced EvalVid. EvalVid combination of NS2 and NS through the expansion of the realization of agents, and the design of three agents, namely MyTrafficTrace, MyUDP and MyUDPSink, the passage of the three agents can be exiled to the real video of the environment simulation. Chih one in the literature only consider Heng Ke, such as network congestion due to such factors as packet loss caused by the impact of video quality, without taking into account the network delay on video quality, especially for real-time video transmission quality. To this end this paper, a video based on the NS network QoS analysis methods, both packet loss and delay to consider the impact on video quality, to achieve the EnhancedEvalVid added.

2 video based on the NS network QoS analysis

2.1 system, the basic framework

In this paper, the structure of the simulation system as shown in Figure 1. Encoder YUV format video will be the source of the request in accordance with the configuration file to encode, resulting in video files. Video streams based on the situation, resulting in network traffic trace file and trace file through the network simulator. Trace generated through simulation results, after analysis of delay and packet loss record record. Finally, the use of these records produce a document video compression injury, decoding can be carried out after the evaluation of the quality.

NS-based network video QoS analysis

2.2 Analysis of the steps

Video transmission due to the simulation needs to be a true simulation of the video stream in the transmission of the network environment, it is necessary to carry out NS-2 to expand and modify, add, video transmission simulation process of the network elements required, including the design of agents. In this paper, the literature, the authors designed three Deputy MyTrafficTrace, MyUDP and MyUDPSink. Acting in accordance with the MyTraffic-Trace Send trace file, at an appropriate time to send packet to the lower UDP, delivery time is based on the rate of video frame to send the document in the trace set. UDP Deputy MyUDP is an extension, it records the time stamp of each packet transmission, packet size and packet id. Myl5DPSink receiving agents, receiving the video packet MyUDP sent, the agent designated to receive the paper record of each packet time stamp, packet size and packet id. In addition, researchers need to verify if the proposed transfer of its strategy, it is necessary to use c + + and OTcl programming on network elements, and its strategy to add the network elements, and then re-compile NS. Upon completion of the NS-2 after the expansion, you can use NS simulation.

2.2.1 generate trace files used in simulation

As a result of video encoding software simulation can not be directly used to produce the trace file. Therefore needs to be generated by video encoder to convert video files trace file, trace file is <packet transmission time, packet marking, packet type, packet size>. The basic principle is to convert video files to read the timestamp and packet size, and the information stored in the trace file. For example, data <O. 066 667, id8, udp, 407>, is the image test sequence mother_daughter. yuv encrypted, the trace file generated by a group of data that sent the size of 0.066 667 s for 407 B, sub-group identified as id8.

2.2.2 Simulation Configuration

(1) In accordance with the requirements of the actual network, the definition of network nodes, configure the network topology to determine the basic characteristics of links, such as delay, bandwidth and so on.

(2) the establishment of agreements agents, including the client device and communication protocol binding model of business will be video streaming and a variety of background binding to the agent in the stream. Configuration parameters of the volume of business model to determine the network distribution business.

(3) set the Trace object. Trace object occurred during the simulation of certain types of events recorded in the trace file. NS-2 through the trace file of the simulation process. Simulation is complete, you can file for trace analysis.

(4) supporting the preparation of other process, setting the end of simulation time, this scripting OTcl completed, and then explain the implementation of NS-2 has been prepared OTcl script, simulation.

2.2.3 After the transmission of video files to restore and decoding

Network Emulator for each packet transmission to produce the records, after the simulation trace files, records all trace simulation data. MyUDPSink agent through the target to generate trace file, its format is <Packet arrival time, packet marking, packet type, packet size>. For example, data <O. 275093, id 8, udp, 407>, said O. In the above example 066 667 s to send the size of 407 B at the moment id8 group O. 275 093 s have been received, if the trace file in the target group did not correspond to the data packet loss. The following two sets of data: <O. 510840, idl3, udp, 102>, <O. 608045, id 14, udp, 306>, respectively corresponding to two consecutive sub-images, as the delay between the two division o. More than 097 205 s inter-frame set by the largest interval, so in real-time video transmission over packet 14 due to delay constraints, will not be able to decode.

Trace file can be seen depending on the target can determine which sub-compressed video to packet loss during transmission, which grouped more than a certain delay due to the restrictions can not be used to decode. Based on this method, you can file from the compressed video transmission of the packet loss and packet delay over the prescribed limit discarded, resulting in the transfer after the new video compression file. Decoder to decode the document will be the reconstruction of the video, which carried out an assessment of the quality.

Case Study 3

Is shown in Figure 2 for examples of this paper, the analysis of the structure of video transmission system diagram, video cameras have a video file, through the timely transmission node S1 to node D1, through the intermediate nodes R1 and R2; node S1 with a CBR traffic generator, but also through intermediate nodes R1 and R2, the node D2 to send, as the context of the impact of streaming video transmission. Bandwidth links in the mark as shown in Figure. The simulation examples to illustrate the main flow of the context of CBR video transmission quality in order to proof the feasibility of this analysis.

NS-based network video QoS analysis

This example uses the image of 250 test sequences mother_daughter. yuv, given the use of JVT reference model JMl. 7 H. Encoder 264 for encoding, resulting in mother_daughter. 264 compressed video files, compressed video programming to read a document, resulting in named mother_daughter. trc the trace file. Figure 2 configured in accordance with the network topology, determine the basic characteristics of links. Video trace file into the part of NS-2 code is as follows:

NS-based network video QoS analysis

NAM dynamic simulation as shown in Figure 3 shows a screenshot, in R1 (Figure 3 on the 2nd node) to produce a large number of packet loss.

NS-based network video QoS analysis

The following are examples of the simulation output generated by the part of the contents of trace files:

NS-based network video QoS analysis

For the meaning of its format, from the first one to the first 12 were as follows: the first one to track the object that the achievement of a specific type of track, there are + (into the queue), one (the queue), r (receive), d (discarded) 4 species; the first two events of that time (unit: seconds); No. 3, No. 4, respectively, to track the source and destination node number (corresponding to Figure 3 in the node); No. 5 Type the name of that package; the first six that packet size; No. 7 is a symbol string is not used in this case; No. 8 is the definition of the flow identifier Ipv6; No. 9,10 package out that the source and destination node address ; No. ll that flow out of the order number; No. 12, said in a simulation package to generate each new identification number only. More than the size of data sets recorded for the 407 B (435 included in the packet header 28 B) the whole process of sub-transmission, and then through the target MyUDPSink agent trace records generated <0.275093, id 8, udp, 407>.

The use of mother_daughter. 264 documents and the target trace file, the preparation of analytical procedures, from mother_daughter. 264 will timeout due to packet loss and packet loss by eliminating, the new video compression files, can be decoded up from the subjective and objective evaluation of video quality.

In Figure 4, respectively PSNR2 and PSNRl whether CBR background showed the existence of two kinds of flow conditions, the PSNR of video sequence comparison shows that the existence of CBR flow deterioration of video quality. Figure 5 from a subjective point of view, comparing the two cases the video quality, by this simple experiment proves that the proposal put forward by this article is totally feasible.

NS-based network video QoS analysis

4 Conclusion

In this paper, selected NS-2 network simulator as a simulation environment, design and realization of a video QoS analysis test platform for all the realization of the process of analysis, through the example of the proposals proved entirely feasible. Network QoS for video of the researchers, can be given practical experiment, analyze and verify the proposed strategy for the effectiveness of video transmission.

Optical Electronics and Display Articles

Can't Find What You're Looking For?

Rating: Not yet rated


No comments posted.