CN102710374A - Speed control method in wireless streaming media transmission - Google Patents

Abstract

The present invention relates to wireless multimedia communication. In order to reduce the transmission quality jitter caused by the unreliable wireless channel when the streaming media is transmitted in a wireless environment, and then improve the transmission quality and the utilization rate of the wireless channel, the technical solution adopted by the present invention is that the rate control in the wireless streaming media transmission method, obtain packet loss rate and service data pass rate parameters through smooth processing in a shorter time interval than traditional rate control methods; adjust the streaming media transmission rate based on the above parameters and the improved increase, multiply, and subtract algorithm, and when the channel is lightly loaded, The variable constant additive factor and constant additive factor determined by the packet loss rate determine the rate increase. When the channel is overloaded, the variable constant multiplicative factor and constant subtractive factor determined by the packet loss rate determine the rate reduction; set the rate Switching threshold, only when the difference between the current predicted rate and the last predicted rate exceeds the set threshold, the rate switching is considered valid. The invention is mainly applied to wireless multimedia communication.

Description

Rate Control Method in Wireless Streaming Media Transmission

technical field

The invention belongs to the technical field of wireless multimedia communication, and relates to a rate control method for guaranteeing service quality of wireless streaming media transmission based on packet loss rate, service data pass rate and an improved AIMD algorithm. Specifically, it relates to a method based on constraint A solver for spatial problems that meets the framework.

Background technique

In recent years, with the rapid development of communication technology, network technology, and video codec technology, the service content of network communication has become more colorful. Multimedia communication combined with information transmission such as voice, video, and data has become an inevitable trend in the development of communication services. Among them, multimedia video communication has become a research hotspot due to its characteristics of continuous transmission, large data volume, and high requirements for real-time performance and reliability.

Wired multimedia communication technology has matured, such as video telephony based on H.261/H.263 standard, video on demand (VoD: Video on Demand), multimedia email, cable digital TV and video surveillance and other technologies have been widely used. However, although wired channels can provide a relatively stable and reliable transmission environment for multimedia communication, the inherent disadvantages of wired transmission, such as high requirements for wiring environment and high cost, limit the further development of multimedia communication. Multimedia communication services based on wireless transmission have attracted people's attention.

In the past ten years, the development of wireless transmission technology has advanced by leaps and bounds, and the transmission bandwidth that wireless networks can provide has also been rapidly improved. The third-generation mobile communication network can provide a rate of 2Mbps for fixed location services. The rate can reach 384Kbps when people are walking, and 128Kbps when driving at a speed of a car, which has reached the bandwidth requirements of multimedia video transmission services. In the fourth generation (4G) wireless communication system, the transmission rate can reach more than 80Mbps; IEEE 802.11 series standards can provide a transmission rate of several megabytes to hundreds of megabytes per second; China recently proposed an ultra-high-speed wireless local area network with independent intellectual property rights The standard even increases the transmission rate of the physical layer to Gbps. The rapid growth of wireless communication bandwidth provides better development opportunities for multimedia communication. However, the wireless channel has a high error rate and time-varying characteristics. The unreliability of the wireless channel will cause severe jitter in the quality of multimedia video, which makes it a huge challenge to ensure the quality of wireless streaming media transmission services.

Based on the above reasons, various rate control methods have been proposed to improve the service quality and transmission efficiency of wireless streaming media transmission. These streaming media rate control methods are widely used in real-time transport protocol (RTP) based streaming media transmission systems. The Real-time Transport Protocol (RTP) is a transport protocol for multimedia data streams on the Internet. The RTP protocol consists of two parts: RTP and RTCP (Real-time Transport Control Protocol). The RTP protocol is located on top of the transport layer protocols TCP and UDP. In order to ensure the real-time transmission, the RTP protocol directly delivers the RTP packet to UDP for transmission after adding the RTP packet header before the payload. RTCP mainly counts various transmission parameters (packet loss rate, delay, etc.) in the transmission process. Traditional rate control algorithms use the transmission parameters provided by RTCP for rate control. According to the rate control principle of the current streaming media rate control algorithm, it can be divided into two categories:

■ Add-increase-multiply-subtract (AIMD) algorithm. AIMD is an algorithm that provides congestion control for the network layer in the TCP/IP transmission model. This algorithm can provide reliable data transmission service for TCP/IP network. It judges the state of the network by comparing the measured packet loss rate with the preset packet loss rate threshold. If the packet loss rate is greater than the packet loss rate threshold, it means that congestion occurs, and the sending rate is reduced according to the preset constant multiplicative reduction factor; otherwise, the sending rate is increased according to the preset constant additive growth factor. The AIMD algorithm does not consider the relationship between the current channel congestion level and the rate adjustment range when performing rate adjustment. It is a linear algorithm and the calculation is very simple. However, using the AIMD algorithm tends to cause saw-tooth fluctuations in the transmission rate, which is not conducive to improving the quality of transmission services.

■TCP Friendly Rate Control (TFRC, TCP Friendly Rate Control) algorithm. The TFRC algorithm adjusts the transmission rate based on the TCP throughput model, as shown in formula (1)

$\mathrm{<span\; class="notranslate">\; SendRate</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&lambda;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1.5</span>}<span\; class="notranslate">\; \&times;</span>\sqrt{\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; 3</span>}}}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; MTU</span>}}{\mathrm{<span\; class="notranslate">\; RTT</span>}<span\; class="notranslate">\; \&times;</span>\sqrt{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}}<span\; class="notranslate">\; \&ap;</span>\frac{\mathrm{<span\; class="notranslate">\; 1.22</span>}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; MTU</span>}}{\mathrm{<span\; class="notranslate">\; RTT</span>}<span\; class="notranslate">\; \&times;</span>\sqrt{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Among them, MTU is the maximum transmission unit; λ is the packet loss rate, RTT is the loopback time, and SendRate is the network throughput; this algorithm can obtain a relatively stable and smooth transmission rate, but the algorithm is based on maximizing the use of network resources. adjustment, ignoring the perceived quality of the video stream.

The above two types of algorithms have a common feature: both types of algorithms use the transmission parameter statistics information provided by the RR (Receiver Report) package in the RTCP protocol to control the streaming media transmission rate. Since the RR packets in the RTCP protocol are sent at intervals of 5s, this means that the streaming media transmission system can only perform transmission rate control every 5s, which is sufficient for wired transmission with relatively stable channels; however, the bandwidth has time-varying Due to the characteristics of wireless transmission, the control of the transmission rate cannot adapt to the change of the channel.

Contents of the invention

The present invention aims to overcome the deficiencies of the prior art, and provides a rate control method for streaming media transmission in a wireless environment, so as to reduce the transmission quality jitter caused by unreliable wireless channels when streaming media is transmitted in a wireless environment, thereby improving the quality of streaming. Quality of service for media transmission and utilization of wireless channels. In order to achieve the above object, the technical solution adopted by the present invention is that the rate control method in the wireless streaming media transmission obtains the packet loss rate and service data pass rate parameters through smooth processing in a shorter time interval than the traditional rate control method; based on the above Parameters and an improved algorithm of adding, multiplying, and subtracting to adjust the streaming media transmission rate. When the channel is lightly loaded, the rate increase is determined by the variable constant additive factor and the constant additive factor determined by the packet loss rate. When the channel is overloaded, the rate increase is determined by the packet loss rate. The variable constant multiplicative factor and constant subtractive factor determined by the rate determine the rate reduction rate; set the rate switch threshold, and only when the difference between the current predicted rate and the last predicted rate exceeds the set threshold, the rate switch is considered valid.

Include the following steps:

(1) Obtain packet loss rate and service data pass rate: at the receiving end of the streaming media transmission system, obtain the packet loss rate and service data pass rate parameters by statistically analyzing the received streaming media data packets;

(2) Predict the new transmission rate: After the receiving end of the streaming media transmission system obtains the packet loss rate and business data pass rate, based on the improved AIMD algorithm, it analyzes the state of the wireless channel at this time and predicts a new suitable wireless channel at this time bandwidth transmission rate;

(3) Determine whether to adjust the transmission rate: After the receiving end of the streaming media transmission system predicts the new transmission rate, it needs to compare whether the difference between the new rate predicted this time and the new rate predicted last time exceeds the set threshold , if the threshold is exceeded, the current rate adjustment is valid; otherwise, the current rate adjustment is invalid.

The steps to obtain the packet loss rate and service data pass rate are as follows:

1) According to the state of the wireless channel, determine the time interval Δt for obtaining the packet loss rate and the service data pass rate;

2) Determine the number N of time intervals participating in the calculation when obtaining the i-th time interval parameter, and the corresponding weights of the i-j time interval participating in the calculation;

3) Statistical streaming media transmission parameters, when the receiving end receives the streaming media data packets, count the parameters in the i-th time interval, including the number of lost packets LPN _i , the total number of sent packets SPN _i , the amount of business data D _i and the exact The interval time T _i ;

4) Calculate the packet loss rate P _i and business data pass rate R _i in the i-th time interval, and calculate the packet loss rate by weighting and smoothing the transmission parameters obtained from the i-N+1 time interval to the i-th time interval and business data pass rate. The time interval Δt in the step of obtaining the packet loss rate and business data pass rate is 2s, and the number N of time intervals involved in the calculation is 2 or 3.

The improved AIMD algorithm contains the following steps:

1) According to the relationship between video quality and packet loss rate at the receiving end of the video transmission system, set the upper and lower thresholds of the packet loss rate in the algorithm as P _{loss_max} and P _{loss_min} , the maximum service data rate R _max and the minimum service data rate R _min ;

2) Compare the relationship between the measured packet loss rate _Pi and the preset packet loss rate threshold, and determine whether the channel state is overloaded, lightly loaded or load balanced;

3) When the channel is overloaded, that is, P _i ≥ P _{loss_max} , calculate the channel overload difference factor m=P _i -P _{loss_max} and the overload rate reduction factor β _k = β _k-1 ×(1-0.6×m) ² , Where k represents the number of consecutive overloads, and the new transmission rate is calculated according to the formula R _new = MAX{β _K × R _i -50, R _min }, where R _i is the service data throughput rate in the i-th time interval, and R _min is The minimum service data rate set by the system;

4) When the channel is lightly loaded, that is, P _i ≤ P _{loss_min} , calculate the channel light load difference factor m=λ×(P _{loss_min} -P _i ) and rate increase α=0.2e ^m ×R ₀ +60, where R ₀ is the preset rate growth factor; calculate the new transmission rate according to the formula R _new ＝MIN{α+R _i , R _max }, where R _i is the service data throughput rate in the i-th time interval, and R _max is the system The maximum service data rate allowed;

5) When the channel load is balanced, there is no need to adjust the rate, and the rate control method is exited.

The value range of the first overload rate reduction factor _β0 in step 3) of the improved AIMD algorithm is [0.85, 0.95]; the value range of the rate increase factor _R0 in step 4) of the improved AIMD algorithm is [80, 150] .

The rate switching threshold is set in the rate control method, and the threshold value R _th is generally set to 40-80 Kb/s.

Technical characteristics and effects of the present invention:

The present invention uses a smaller channel state acquisition time interval to analyze the channel state, so that the method can quickly reflect the change of the wireless channel. The invention predicts a suitable transmission rate based on the improved AIMD algorithm. Compared with the traditional AIMD algorithm, the improved algorithm increases the impact of the packet loss rate on the adjustment range of the transmission rate, and can make the rate prediction closer to the wireless channel bandwidth. The invention fully considers the characteristics of the wireless channel, and can be used as an effective method for improving the quality of streaming media transmission in a wireless environment.

The beneficial effect of the present invention also lies in:

■ Quickly reflect changes in wireless channels. The wireless channel has characteristics such as fading and time-varying, and the channel state changes relatively quickly. The present invention obtains the packet loss rate and the service data pass rate based on a self-defined smaller time interval, and at the same time smooths important parameters, which conforms to the time-varying characteristics of the wireless channel, making the adjustment of the rate more timely;

■ Reduced transmission rate jitter. Through the improved AIMD rate self-adaptive adjustment algorithm, the present invention makes full use of the two important parameters of packet loss rate and service data pass rate to adjust the transmission rate, making the rate adjustment more accurate and reducing sawtooth fluctuations;

■ Reduce system overhead. The present invention can reduce unnecessary overhead of the system by adding a judgment mechanism before the transmission rate adjustment feedback;

■Improve the service quality of wireless streaming media transmission and the utilization rate of wireless channel. The invention reduces the packet loss rate and the jitter of the transmission rate by controlling the rate in the streaming media transmission process, thereby improving the transmission service quality and channel utilization rate.

Description of drawings

Figure 1 shows the overall flow chart of the rate control method.

Fig. 2 is a flowchart of a method for obtaining packet loss rate and service data pass rate.

Figure 3 shows the general flow chart of the improved AIMD algorithm.

Figure 4 is a flow chart of rate adjustment in the case of channel overload in the improved AIMD algorithm.

Figure 5 is a flow chart of rate adjustment in the case of channel light load in the improved AIMD algorithm.

Fig. 6 is a schematic diagram of the structural composition of the wireless real-time video transmission system.

Fig. 7 is a peak signal-to-noise ratio diagram of the system performance test results of the embodiment.

Figure 8 is a diagram of the packet loss rate and service data pass rate of the system performance test results of the embodiment.

Detailed ways

Based on the above considerations, the present invention proposes a rate control method for wireless streaming media transmission aiming at improving the service quality and channel utilization rate of wireless streaming media transmission. This method is based on the streaming media transmission system, and at the receiving end of the system, the parameters such as the packet loss rate and the business data passing rate are counted at a shorter time interval than the traditional rate control method. Based on the above two parameters, the streaming media is predicted by the improved AIMD algorithm. The service data transmission rate is fed back to the sender by the TCP protocol after the rate switching is determined to be valid. Compared with the traditional streaming media rate control method, the invention can make the streaming media transmission rate adapt to the bandwidth of the wireless channel more accurately, thereby ensuring the service quality of the streaming media transmission and improving the utilization rate of the wireless channel.

In order to achieve the above object, the present invention provides a rate control method in wireless streaming media transmission, which is characterized in that: the method obtains the packet loss rate and service data through smooth processing in a shorter time interval than the traditional rate control method Rate and other parameters; adjust the streaming media transmission rate based on the above parameters and the improved increase, increase, multiply and subtract algorithm. When the channel is lightly loaded, the variable constant additive factor and the constant additive factor determined by the packet loss rate determine the rate increase. When overloaded, the variable constant multiplicative factor and the constant subtractive factor determined by the packet loss rate determine the rate reduction rate; finally, the present invention sets the rate switching threshold, and only when the difference between the current predicted rate and the last predicted rate exceeds the set The rate switching is considered valid only when the threshold is set.

The present invention provides a rate control method in wireless streaming media transmission, characterized in that: the method includes the following steps:

(1) Obtain the packet loss rate and service data pass rate. At the receiving end of the streaming media transmission system, parameters such as packet loss rate and service data passing rate are obtained by statistically analyzing the received streaming media data packets;

(2) Predict the new transmission rate. After the receiving end of the streaming media transmission system obtains the packet loss rate and the service data pass rate, based on the improved AIMD algorithm, it analyzes the state of the wireless channel at this time and predicts a new transmission rate suitable for the wireless channel bandwidth at this time;

(3) Determine whether to adjust the transmission rate. After the receiving end of the streaming media transmission system predicts the new transmission rate, it needs to compare whether the difference between the new rate predicted this time and the new rate predicted last time exceeds the set threshold. If it exceeds the threshold, the current rate The adjustment is valid, otherwise, the current rate adjustment is invalid.

The present invention provides a rate control method in wireless streaming media transmission, which is characterized in that: the steps of obtaining the packet loss rate and service data pass rate are as follows:

1) According to the state of the wireless channel, determine the time interval Δt for obtaining the packet loss rate and the service data pass rate;

2) Determine the number N of time intervals participating in the calculation when obtaining the packet loss rate and business data pass rate of the i-th time interval, and the corresponding weights α _j and μ _j of the ij-th time interval participating in the calculation, where 0≤j≤ N;

3) Statistical streaming media transmission parameters. While receiving streaming media data packets at the receiving end, count the parameters in the i-th time interval, including the number of lost packets LPN _i , the total number of sent packets SPN _i , the amount of business data D _i and the accurate interval time T _i ;

4) Calculate the packet loss rate and business data pass rate of the i-th time interval. The packet loss rate and the service data pass rate are calculated by performing weighted smoothing on the transmission parameters obtained from the i-N+1 time interval to the i-th time interval.

The rate control method in the wireless streaming media transmission proposed by the present invention is characterized in that: the time interval Δt in the step of obtaining the packet loss rate and the service data pass rate is generally about 2s, and the number N of time intervals participating in the calculation is 2 or 3.

The rate control method in the wireless streaming media transmission proposed by the present invention is characterized in that: the improved AIMD algorithm comprises the following steps:

1) According to the relationship between video quality and packet loss rate at the receiving end of the video transmission system, set the upper and lower thresholds of the packet loss rate in the algorithm as P _{loss_max} and P _{loss_min} , the maximum service data rate R _max and the minimum service data rate R _min ;

2) Compare the relationship between the measured packet loss rate _Pi and the preset packet loss rate threshold, and determine whether the channel state is overloaded, lightly loaded or load balanced;

3) When the channel is overloaded, that is, P _i ≥ P _{loss_max} , calculate the channel overload difference factor m=P _i -P _{loss_max} and the overload rate reduction factor β _k = β _k-1 ×(1-0.6×m) ² , Where k represents the number of consecutive overloads, and the new transmission rate is calculated according to the formula R _new = MAX{β _K × R _i -50, R _min }, where R _i is the service data throughput rate in the i-th time interval, and R _min is The minimum service data rate set by the system;

4) When the channel is lightly loaded, that is, P _i ≤ P _{loss_min} , calculate the channel light load difference factor m=λ×(P _{oss_min} -P _i ) and rate increase α=0.2e ^m ×R ₀ +60, where R ₀ is the preset rate growth factor; calculate the new transmission rate according to the formula R _new ＝MIN{α+R _i , R _max }, where R _i is the service data throughput rate in the i-th time interval, and R _max is the system The maximum service data rate allowed;

5) When the channel load is balanced, there is no need to adjust the rate, and the rate control method is exited.

The rate control method in the wireless streaming media transmission proposed by the present invention is characterized in that: the improved AIMD algorithm realizes step 3) in the value range of first overload rate reduction factor β ₀ is [0.85,0.95]; Improved AIMD algorithm The value range of the rate increase factor _R0 in step 4) is [80, 150].

The rate control method in the wireless streaming media transmission proposed by the present invention is characterized in that a rate switching threshold is set in the rate control method, and the threshold value R _th is generally set at 40-80 Kb/s.

In order to make the purpose, technical solution features and advantages of the present invention clearer, the present invention will be described in detail below in conjunction with the accompanying drawings.

The invention is a rate control method for streaming media transmission in a wireless channel environment. Streaming media transmission, especially real-time video transmission, has the characteristics of continuous transmission, large amount of data, and high requirements for real-time performance and reliability. If the channel is unreliable or congested, the service quality of streaming media transmission will be seriously affected. In the case of a wired network, the main reason affecting the quality of streaming media transmission is the congestion of the network. However, in a wireless environment, the main factor affecting the quality of streaming media transmission is the time-varying and unreliable bandwidth of the wireless channel. Therefore, the present invention mainly aims at the streaming media transmission in the wireless environment, and judges the state of the wireless channel by analyzing the transmission situation for a period of time, and finally predicts an appropriate transmission rate. The invention can reduce the packet loss rate in the streaming media transmission process, thereby improving the service quality and wireless channel utilization rate of the wireless streaming media transmission.

The invention adopts a shorter time interval than the traditional stream media rate control method to obtain the packet loss rate and the service data pass rate, and uses the improved AIMD algorithm to predict the new transmission rate. This enables the method to quickly reflect the change of the wireless channel and accurately predict the new transmission rate, thereby improving the service quality of wireless streaming media transmission and the utilization rate of the wireless channel.

Before using the present invention, it is necessary to determine the preset parameters of the present invention according to the actual streaming media application scenario and the wireless environment, and these parameter values have a great influence on the performance of the rate control method provided by the present invention.

■ Maximum service data rate R _max and minimum service data rate R _mim . These two parameters are the maximum value and the minimum value of the service data rate adjustment respectively. Their values are determined according to the actual requirements of the streaming media service on the service data throughput.

■The maximum packet loss rate threshold P _{loss_max} and the minimum packet loss rate threshold P _{loss_min} . The maximum packet loss rate threshold indicates the tolerance of the streaming media transmission system to the packet loss rate, and the minimum packet loss rate threshold indicates the tolerance of the streaming media transmission system to the channel light load.

■ Time interval Δt for rate control. This value indicates that in the process of streaming media transmission, the streaming media transmission rate is adjusted according to the rate control algorithm of the present invention at intervals of Δt. If the value is too large, it cannot reflect the time-varying nature of the wireless channel; if the value is too small, the rate jitter will be large.

Each step of the method of the present invention will be described in detail below.

Fig. 1 is the general flowchart of the method of the present invention. It can be seen from the figure that the method of the present invention mainly includes three parts: acquisition of packet loss rate and service data pass rate, rate adjustment of the improved AIMD algorithm, and system rate switching judgment. Based on the time interval Δt of the rate control being set, the method of the present invention obtains the streaming media transmission parameters within the time period Δt, and uses the improved AIMD algorithm to predict the new transmission rate at the last moment of Δt, and at the same time checks the effectiveness of the system rate switching determination. The rate control method of the present invention always runs cyclically with Δt as the time interval. In order to make the detailed process described below clearer, the ith time interval is taken as an example for illustration.

The detailed flow process of the present invention is as follows:

(1) Packet loss rate and service data pass rate. While the receiving end is receiving the streaming media data packet, the transmission information of the streaming media data packet in the i-th time interval (including the total number of data packets sent by the sending end, the number of data packets received by the receiving end, and the total number of data packets received by the receiving end business data volume and accurate time interval value) to calculate the packet loss rate P _i and business data pass rate R _i in the i-th time interval;

(2) Predict the new transmission rate. After obtaining the packet loss rate and service data pass rate, the receiving end compares the packet loss rate P _i with the preset packet loss rate threshold to know the current channel status: overload, light load and balance. According to the three channel states, packet loss rate, and service data pass rate, the improved AIMD algorithm can accurately predict the new rate;

(3) Determine whether to adjust the transmission rate. When the new transmission rate R _new is obtained, compare this value with the last rate prediction value R _{new_last} , and only when |R _new - R _{new_last} | > R _th , the streaming media system is notified to change the transmission rate, where R _th is a preset rate adjustment threshold.

The processing flow of obtaining the packet loss rate and service data pass rate in Figure 1 is shown in Figure 2. The receiving end obtains the sequence number of the current data packet, the video frame to which the current data packet belongs, and the length of the packet payload data through the received streaming media data packet and other information to calculate the parameters required for rate adaptive adjustment.

The process of obtaining packet loss rate and business data pass rate is as follows:

(1) Initialize the counter first. It includes initializing the video frame number counter F_cnt, the received stream media data packet counter P_cnt and the service data amount counter D _i . Since the data volume of the video frame is relatively large, the sender will cut it into several data packets for transmission. Therefore, every time a streaming media data packet is received, it is first necessary to judge whether the data packet is the initial data packet of a video frame according to the flag bit preset by the sending end, if so, add 1 to the frame counter F_cnt, otherwise it remains unchanged; Then the payload service data length of the data packet is obtained, and the length is added to the service data amount counter D _i ; finally, the packet counter P_cnt is incremented by 1 accordingly.

(2) Determine the counting start parameters. When the counter P_cnt of the received streaming media data packet is 1, the channel prediction acquisition of a new time period starts, and the sequence number psn1 of the data packet is obtained (in order to check whether the streaming media transmission is lost, a 16-bit value needs to be added to each data packet serial number) and the current time time1;

(3) Determine the counting termination parameter. In order to adapt to the rapid change of the wireless channel, the system is set that when the number of received frames reaches 30 frames, the time period ends, and the sequence number psn2 of the data packet and the current time time2 are obtained at this time. The video frame rate of the system is 15 frames/s, so each time interval is about Δt=2s;

(4) Calculate the exact value of the time interval T _i =time2-time1, and the total number of packets SPN _i =psn2-psn1+1 sent by the sender within the time T _i , and the number of lost packets LPN _i =SPN _i -P_cnt. Since the packet sequence number in the data packet header occupies 16 bits, the maximum sequence number is 65535, and the sequence number overflows when the system is running, that is, the number of sent packets SPN _i <0 occurs, and at this time SPN _i =psn2-psn1+1+65536.

(5) Calculate the packet loss rate P _i . Since the prediction period Δt is small, and the packet loss rate reflects the congestion state of the channel, which is related to the channel congestion state at the previous moment, the weighted value of the statistical results of various data packets in the previous N time intervals is used for calculation, and the N value depends on in the channel environment. Calculate the packet loss rate according to the formula (2), where when N=3, the weighted value α is: α ₀ =0.5, α ₁ =0.3, α ₂ =0.2; when N=2, the weighted value α is The values are: α ₀ =0.7, α ₁ =0.3; LPN _ij represents the total number of lost packets collected in the ij-th time interval, and SPN _ij represents the total number of data packets sent by the sender in the ij-th time interval.

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; LPN</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; j</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; SPNs</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; ,</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

(6) Calculate the business data pass rate R _i . Calculate the service data pass rate according to formula (3), where when N=3, the weighted value μ ₀ =0.7, μ ₁ =0.2, μ ₂ =0.1; when N=2, the weighted value μ ₀ =0.8, μ ₁ = 0.2; D _ij represents the amount of business data received in the ij time interval, and T _ij is the exact time value of the ij time interval;

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\frac{\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; j</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; ,</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

(7) Reset the various counters used above, and start the channel prediction calculation for the next time interval.

The overall flow chart of the improved AIMD algorithm in Figure 1 is shown in Figure 3. The traditional AIMD algorithm does not consider the influence of real-time packet loss rate parameters. When the channel is overloaded, the transmission rate is reduced by multiplying the current rate by a constant factor. When the channel is lightly loaded, the transmission rate is increased by adding a constant factor to the current transmission rate. The sawtooth fluctuation of the rate. The present invention considers the packet loss rate as an important parameter when adjusting the rate, and the greater the difference between the packet loss rate and the threshold value, the greater the range of rate adjustment.

The steps of the improved AIMD algorithm are as follows:

(1) Setting the maximum packet loss rate threshold P _{loss_max} and the minimum packet loss rate threshold P _{loss_min} allowed by the streaming media transmission system;

(2) Comparing the packet loss rate _Pi with a preset packet loss rate threshold. If the packet loss rate P _i is greater than the maximum packet loss rate threshold P _{loss_max} , it means that the current channel is overloaded, and the rate needs to be reduced to improve the video transmission service quality;

(3) If the packet loss rate P _i is less than the minimum packet loss rate threshold P _{loss_min} , it means that the current channel is lightly loaded, and the rate needs to be increased to improve the utilization rate of the wireless channel;

(4) If the packet loss rate P _i is greater than the minimum packet loss rate threshold P _{loss_min} and smaller than the maximum packet loss rate threshold P _{loss_max} , it means that the channel load is balanced, no rate adjustment is required, and the rate control algorithm is exited.

The processing flow when the channel is overloaded in Figure 2 is shown in Figure 4. The improved AIMD algorithm determines the new transmission rate by comprehensively considering the degree of channel overload and the current service data throughput rate of the channel. The new rate determination process when overloaded is as follows:

(1) Initialize the first time channel overload rate reduction factor β ₀ , which affects the rate reduction range, which is 0.93 in the present invention;

(2) Judge the times k of continuous channel overloading. When the channel is continuously overloaded, it means that the channel has entered a relatively large fading, and it is necessary to increase the rate reduction range relatively.

(3) Calculate the difference factor m=P _i -P _{loss_max} for this channel overload, where P _i is the measured packet loss rate, and P _{loss_max} is the maximum packet loss rate preset by the system;

(4) Calculating the channel overload rate reduction factor β _k = β _k-1 × (1-0.6 × m) ² ;

(5) Calculate the new rate R _new ＝MAX{β _K ×R _i -50, R _min } generated by this rate adaptive adjustment, where R _i is the service data throughput rate, and R _min is the system preset minimum service data transmission rate;

The processing flow when the channel is lightly loaded in Figure 2 is shown in Figure 5. The rate adaptive control module determines the new transmission rate by comprehensively considering the degree of channel light load and the current service data throughput rate of the channel. The new rate determination process at light load is as follows:

(1) Initialize the rate increase factor R ₀ , which affects the range of rate increase, and the value is 100 in the system;

(2) The recovery rate is reduced by a factor β ₀ to prepare for the next channel overload;

(3) Calculate the packet loss rate difference factor m=λ×(P _{loss_min} -P _i ) under light load of the channel this time, where λ is the magnification factor of the packet loss rate difference;

(4) Calculate the rate increase under light channel load this time α = 0.2e ^m × R ₀ +60, where R ₀ is the rate increase factor;

(5) Calculate the new rate R _new = MIN{α+R _i , R _max } generated by the rate adaptive adjustment this time, where R _i is the service data throughput rate in the i-th time interval, and R _max is the maximum service data rate ;

The wireless streaming media rate control method provided by the present invention can be used in various scenarios, including wireless video monitoring, wireless video conferencing, wireless video telephony, streaming media downloading and playing, and the like. The method provided by the invention can effectively improve the service quality of these applications.

In the following, real-time video applications with higher requirements on service quality are taken as the embodiment of the present invention. The streaming media rate control method is used in an end-to-end wireless real-time video transmission system, thereby ensuring the service quality of real-time video during transmission.

FIG. 6 is a system block diagram of a specific real-time solution of the present invention, which is an end-to-end wireless video real-time transmission system based on an embedded Linux platform. The system uses the 802.11b/g standard Ad-Hoc (wireless self-organizing network) mode as the wireless environment for real-time video transmission. Because the 802.11b/g standard physical layer rate can reach up to 54Mb/s, it can meet the bandwidth requirements of real-time video transmission. The system consists of a video sending end and a video receiving end. The video sending end mainly includes a video capture encoding module and a video data sending module. The video receiving end mainly includes a data receiving module, a video decoding display module and a rate control module. Each module of the system and its characteristics are described in detail below.

Video sender:

(1) Video capture and encoding module. This module acquires real-time video that meets the requirements by configuring parameters such as the frame rate and resolution of camera video collection. In order to reduce the bandwidth required for transmission, the encoding module encodes it in H264 video format. When the system is running, the main parameters of the acquisition and encoding module are as follows:

■Video resolution: 640x400, remains unchanged during transmission;

■Video frame rate: 15 frames/s, which remains unchanged during transmission;

■Encoder output service data rate: initialized to 2048Kb/s, then adjusted by the rate control module;

(2) Video data sending module. In order to ensure the real-time transmission of video data, we use RTP as the transmission protocol of video data stream. Using RTP as the video data transmission protocol can use the present invention more conveniently, because the RTP protocol stipulates the information carried by the RTP header when the streaming media information is transmitted, including the source of the RTP data packet, the serial number and the time stamp of the data packet. According to the source of the RTP data packet, the streaming media transmission system can receive multiple media sources at the same time at the receiving end; according to the sequence number of the RTP data packet, the packet loss rate of the streaming media transmission system within a certain period of time can be calculated. By obtaining the length of the service data of the RTP data packet, the service data throughput rate of the streaming media transmission system within a period of time can be calculated. In this module, the system cuts and packages a frame of video data according to RTP, and then delivers it to UDP for transmission.

Video receiver:

(1) Data receiving module. The function of this module is to receive RTP data packets, and then unpack the RTP data packets, and combine the RTP data packet payloads of the same frame into a video frame.

(2) Video decoding display module. After the data receiving module synthesizes the received RTP data into a compressed video frame, the decoding module decodes it in H264 video format. Since the decoded video data is in YUV420 format, it needs to be converted to RGB format and displayed by LCD.

(3) Speed control module. This module is the embodiment of the method provided by the present invention in this embodiment, and is an important module for guaranteeing the service quality of real-time video transmission of the system. This module is mainly composed of three sub-modules, including the packet loss rate and business data pass rate acquisition sub-module, the improved AIMD rate adjustment sub-module and the system rate switching judgment sub-module. The descriptions of the three sub-modules are as follows:

■Packet loss rate and business data pass rate acquisition sub-module. The sub-module counts the transmission parameters according to the method provided by the present invention while the data receiving module receives the RTP data packets, so as to obtain the packet loss rate and the service data pass rate. The system sets the cycle of obtaining the above parameters as the time when the receiving end receives 30 frames of video. Since the system video frame rate remains unchanged at 15 frames/s, the time is about 2s. The value of parameter when this submodule calculates packet loss rate and service data passing rate according to formula (2) and (3) provided by the present invention is as follows:

参与计算的时间间隔个数N＝3；

The number of time intervals involved in the calculation N=3;

Packet loss rate smoothing factor α ₀ =0.5, α ₁ =0.3, α ₂ =0.2;

μ₀＝0.7，μ₁＝0.2，μ₂＝0.1。

μ ₀ =0.7, μ ₁ =0.2, μ ₂ =0.1.

■Improved AIMD rate adjustment submodule. This sub-module adjusts the transmission rate according to the improved AIMD algorithm provided by the present invention after obtaining the packet loss rate and the service data passing rate, and the values of various parameters are as follows:

丢包率上下门限值，P_{loss_max}＝0.03，P_{loss_min}＝0.008；

Upper and lower thresholds of packet loss rate, P _{loss_max} = 0.03, P _{loss_min} = 0.008;

最大业务数据速率R_max＝2048kb/s，最小业务数据速率R_min＝256kb/s；

The maximum service data rate R _max =2048kb/s, the minimum service data rate R _min =256kb/s;

The first channel overload rate reduction factor β ₀ =0.93;

Rate increase factor R ₀ =100.

■The system rate switching judgment sub-module. This module is mainly to prevent frequent switching of the system from increasing system overhead. According to the method for judging system rate switching provided by the present invention, the threshold R _th for judging is set to 60Kb/s.

As shown in Figure 6, the system has two wireless transmission links. The first is a video data transmission link based on RTP/UDP/IP, which is mainly responsible for real-time video transmission; the second is a feedback control link based on TCP/IP, which is mainly responsible for the reliable transmission of the rate adjustment information of the rate control module .

In order to evaluate the performance of the above system, parameters such as packet loss rate, business data pass rate, and PSNR (peak signal-to-noise ratio) of video transmission were counted during the video transmission process of the system. The test environment for system performance is a corridor with many different wireless access points, the interference between different wireless signals is relatively large, and the distance between the sending end system and the receiving end system is 42m. The results of the performance tests are as follows.

As shown in Figure 7, it is the PSNR value of the video data after H264 decoding at the sending end and H264 decoding at the receiving end of the calculation system before H264 compression. Therefore, the loss of video quality shown in Figure 7 consists of two parts: transmission loss and H264 codec loss.

The PSNR value of video transmission has a certain relationship with people's evaluation of video perception quality. The PSNR value in Figure 7 is segmented and counted as follows:

PSNR值大于37，视频等级为5，视频质量非常好，占总数88.312％；

The PSNR value is greater than 37, the video level is 5, and the video quality is very good, accounting for 88.312% of the total;

PSNR值位于区间(31，37]，视频等级为4，视频质量好，占总数9.5238％；

The PSNR value is in the interval (31, 37], the video level is 4, and the video quality is good, accounting for 9.5238% of the total;

PSNR值位于区间(25，31]，视频等级为3，视频质量一般，占总数0％；

The PSNR value is in the interval (25, 31], the video level is 3, and the video quality is average, accounting for 0% of the total;

The PSNR value is in the interval (20, 25], the video level is 2, and the video quality is poor, accounting for 0.8658% of the total;

PSNR值小于20，视频等级为4，视频质量差，占总数1.2987％；

The PSNR value is less than 20, the video level is 4, and the video quality is poor, accounting for 1.2987% of the total;

From the statistical results, it can be seen that the system can provide high video transmission service quality, 97.8358% of the cases can provide good video quality, only when the wireless channel fading is very large, it will temporarily cause the video quality to decrease.

FIG. 8 is a parameter diagram of the service data rate and packet loss rate counted when the system performs video transmission. It can be seen from the figure that in most of the time, the packet loss rate of the system is controlled below 2%. In relatively large fading, through the rate control method of the present invention, the service data rate will immediately respond to the change of the wireless channel and reduce the transmission rate, thereby ensuring a lower packet loss rate.

Claims (7)

1. a rate control method in wireless streaming media transmission, it is characterized in that, in the time interval shorter than traditional rate control method, obtain packet loss rate and business data passing rate parameter by smooth processing; Based on above-mentioned parameter and improved plus The multiplication and subtraction algorithm adjusts the streaming media transmission rate. When the channel is lightly loaded, the variable constant additive factor and the constant additive factor determined by the packet loss rate determine the rate increase. When the channel is overloaded, the variable constant determined by the packet loss rate The multiplicative factor and the constant subtractive factor determine the rate reduction rate; set the rate switching threshold, and only when the difference between the current predicted rate and the last predicted rate exceeds the set threshold, the rate switching is considered valid. 2. the rate control method in the wireless streaming media transmission as claimed in claim 1, is characterized in that, comprises the following concrete steps: (1) Obtain packet loss rate and service data pass rate: at the receiving end of the streaming media transmission system, obtain the packet loss rate and service data pass rate parameters by statistically analyzing the received streaming media data packets; (2) Predict the new transmission rate: After the receiving end of the streaming media transmission system obtains the packet loss rate and business data pass rate, based on the improved AIMD algorithm, it analyzes the state of the wireless channel at this time and predicts a new suitable wireless channel at this time bandwidth transmission rate; (3) Determine whether to adjust the transmission rate: After the receiving end of the streaming media transmission system predicts the new transmission rate, it needs to compare whether the difference between the new rate predicted this time and the new rate predicted last time exceeds the set threshold , if the threshold is exceeded, the current rate adjustment is valid; otherwise, the current rate adjustment is invalid. 3. the rate control method in the wireless streaming media transmission as claimed in claim 1, is characterized in that, obtains packet loss rate and service data passing rate step as follows: 1) According to the state of the wireless channel, determine the time interval Δt for obtaining the packet loss rate and the service data pass rate; 2) Determine the number N of time intervals participating in the calculation when obtaining the i-th time interval parameter, and the corresponding weights of the i-j time interval participating in the calculation; 3) Statistical streaming media transmission parameters, when the receiving end receives the streaming media data packets, count the parameters in the i-th time interval, including the number of lost packets LPN _i , the total number of sent packets SPN _i , the amount of business data D _i and the exact The interval time T _i ; 4) Calculate the packet loss rate P _i and business data pass rate R _i in the i-th time interval, and calculate the packet loss rate by weighting and smoothing the transmission parameters obtained from the i-N+1 time interval to the i-th time interval and business data pass rate. 4. The rate control method in wireless streaming media transmission as claimed in claim 1, characterized in that the time interval Δt in the step of obtaining the packet loss rate and the service data passing rate is 2s, and the number N of time intervals participating in the calculation is valued 2 or 3. 5. the rate control method in the wireless streaming media transmission as claimed in claim 1, is characterized in that, improved AIMD algorithm comprises the following steps: 1) According to the relationship between video quality and packet loss rate at the receiving end of the video transmission system, set the upper and lower thresholds of the packet loss rate in the algorithm as P _{loss_max} and P _{loss_min} , the maximum service data rate R _max and the minimum service data rate R _min ; 2) Compare the relationship between the measured packet loss rate _Pi and the preset packet loss rate threshold, and determine whether the channel state is overloaded, lightly loaded or load balanced; 3) When the channel is overloaded, that is, P _i ≥ P _{loss_max} , calculate the channel overload difference factor m=P _i -P _{loss_max} and the overload rate reduction factor β _k = β _k-1 ×(1-0.6×m) ² , Where k represents the number of consecutive overloads, and the new transmission rate is calculated according to the formula R _new = MAX{β _K × R _i -50, R _min }, where R _i is the service data throughput rate in the i-th time interval, and R _min is The minimum service data rate set by the system; 4) When the channel is lightly loaded, that is, P _i ≤ P _{loss_min} , calculate the channel light load difference factor m=λ×(P _{loss_min} -P _i ) and rate increase α=0.2e ^m ×R ₀ +60, where R ₀ is the preset rate growth factor; calculate the new transmission rate according to the formula R _new ＝MIN{α+R _i , R _max }, where R _i is the service data throughput rate in the i-th time interval, and R _max is the system The maximum service data rate allowed; 5) When the channel load is balanced, there is no need to adjust the rate, and the rate control method is exited. 6. the rate control method in the wireless stream media transmission as claimed in claim 1 is characterized in that, the value range of first overload rate reduction factor β ₀ in the improved AIMD algorithm realization step 3) is [0.85,0.95]; The improved AIMD algorithm realizes the value range of the rate increase factor R ₀ in step 4) is [80, 150]. 7. The rate control method in wireless streaming media transmission as claimed in claim 1, wherein a rate switching threshold is set in the rate control method, and the threshold value R _th is generally set to 40-80 Kb/s.

Images