The present invention relates to the wideband code division multiple access (WCDMA) communication system in the field of wireless telecommunication, specifically, relates to a method for scheduling packet in a WCDMA communication system.
Wideband code division multiple access is a scheme of wireless transmission technique proposed by the third generation mobile telecommunication system standard organization. The well-known time division multiple access (TDMA) and frequency division multiple access (FDMA) techniques are mainly used for transmitting voice and data with low rate, so they have certain limitations. The code division multiple access (CDMA) modulation technique is a new kind of telecommunication techniques used in the situation when there are a large number of system users.
The third generation telecommunication system can provide mobile users with multi-media service similar to that of the fixed network, and supports worldwide roam with good service quality. Its biggest character is that it supports wireless services with high transmission rate, such as surfing and browsing internet in wireless condition, downloading, video with high quality, audio stream service, FTP service, visible telephone and e-mail. In the third generation partnership project (3GPP) standard protocol, the R99 version can support air transmission rate as high as 2 Mbps, in the enhanced R5 version, the high speed downlink packet access HSDPA system employs the adaptive modulation code AMC and the hybrid automatic repeat request technique HARQ, so that the third generation mobile telecommunication system can support packet service with the air transmission rate as high as 12 Mbps. Because the packet service to be transmitted has very high rate, the ways of properly scheduling the packets to the air link for transmitting is an important factor in influencing the final transmission efficiency. The indexes for evaluating the packet scheduling algorithm mainly are the peak transmission rate, the extent to which the service quality QoS is met, and the fairness level of treating the packets of respective users.
In wire data telecommunication network, there has already been the technique of packet scheduling, the purpose of which is mainly to improve the transmission efficiency of network, for example, US patent application 20030103453 “System and method for managing flow bandwidth utilization in a packet communication environment” proposed a Time-Division-Queue Rate Control Scheme (TDQ-RCS) to resolve the problem of packet scheduling in wire data telecommunication. However, since in the wire telecommunication system, the analysis on packet scheduling, generally speaking, is based on two kinds of channel states, which are the availability of channel and the non-availability of channel, rather than the channel quality, while there may be many kinds of channel states in the wireless telecommunication system, this invention can not solve the problem of packet scheduling in wireless telecommunication system. The scheduling method used in wireless telecommunication is much more complicated than that in wire data telecommunication.
In order to meet the need of packet scheduling in wireless telecommunication, it is necessary to add the influence of channel condition on scheduling to the original packet scheduling method in wire data telecommunication. For example, US patent application 20020044527 “Channel efficiency based packet scheduling for interactive data in cellular networks” proposed a packet scheduling method used in GPRS system, which considered the influence of channel efficiency parameter on packet scheduling weight in connection with the commonly used weighted fair queuing (WFQ) scheduling method. The channel efficiency is the reciprocal of the maximum transmission rate permitted in the wireless condition, which represents the channel quality corresponding to each user, the better the channel quality is, the more the resource assigned to the packet is. This method is very efficient for the packet service without time delay sensitivity, but as for packet service sensitive to transmission time delay, it can not meet the requirements of service quality QoS. Because this method takes channel quality as the chief judging principle, the packet service with good channel quality takes less time in waiting for transmitting than the packet service with poor channel quality does. Therefore, the transmission of the service sensitive to the transmission time delay will be delayed due to its poor channel quality, and its QoS requirements will not be met.
There are also corresponding solving methods related to the specific packet services with time delay sensitivity. For example, US patent application 20030101274 “packet transmission scheduling technique” described a method for scheduling the real time packet service in WCDMA high speed downlink packet access HSDPA system, which judges the weight of packet by the maximum time delay threshold determined by the priority, quantity, time delay and service quality QoS of the packet on the basis of states of the wireless channel. The closer the time delay of data approximates to the maximum time delay threshold, or when the time delay is long enough, the higher transmission priority the data has. However, as for the real time packet service whose data delay excess the maximum time delay threshold, its corresponding data packet will be discarded. The discarding of packet is unavoidable, and with the existing technique, the lost ratio of packet can not be further decreased, and there is no related solution provided for the users losing packet. As a result, the lost ratio of packet of the users in bad transmission condition remains high, and the communication quality and satisfaction of users may be influenced; Compared with wire transmission, the lost packet in wireless transmission often occurs in burst, so special measures are needed to deal with the packet lost.
The above methods only consider some of the characters of wireless telecommunication while neglect some others. For example, some of the real time packet services, such as the packet audio and video service, are quite sensitive to time delay jitter, which will greatly affect the quality of telecommunication. The common solution to this problem is to set data buffer with enough length to compensate the shortcomings of time delay jitter, as in US paten application 20030112796 “Voice and data exchange over a packet based network with fax relay spoofing” and US patent application 20030026275 “Dynamic Jitter buffering for voice-over-IP and other packet-based communication systems”. However, in order to control the big time delay jitter, big data buffer has to be set, which will result in the increasing transmission time delay of the date, which when excesses a certain limitation, the requirement of service quality of wireless telecommunication can not be met. The existing technique does not consider the treatment toward the time delay jitter in packet scheduling under the condition of meeting all the requirements of service quality QoS, therefore the communication quality of some services is destroyed.
The technical problem to be solved in the present invention is to provide a method for scheduling packet in a wireless telecommunication system, for the purposes of decreasing the lost ratio of packet and controlling the time delay jitter, therefore optimizing the packet time delay, throughput, time delay jitter and lost ratio of packet under the condition of meeting the QoS requirements of all services.
The method for scheduling packet in wireless telecommunication system described in the present invention divides the user packet queues to be transmitted into the user packet queues with lost packet and user packet queues without lost packet; for the user packet queues with lost packet, if a real time lost ratio of packet for the user excesses a predetermined lost ratio threshold of packet, terminates a connection to the user; if the real time lost ratio of packet for the user does not excess the predetermined lost ratio threshold of packet, schedules the user packet queues according to a volume of the lost ratio of packet; for the user packet queues without lost packet, schedules according to packet lengths, channel quality states, time delays and time delay jitters.
The present invention, through giving priority to scheduling the users with high lost ratio of packet under the condition of existing a certain range of packet lost, decreases the lost ratio of packet, especially those whose users are in bad transmission condition, thereby ensures the comparative fairness of transmission in unfair transmission conditions. When the lost ratio of packet excesses the threshold, the connections to the users will be terminated temporarily waiting for the improvement of transmission condition. The present invention takes the requirements of packet service for users sensitive to the time delay jitter into full consideration, it controls the time delay jitter to maintain invariable, therefore improves the telecommunication quality of those users.
Next, the technical scheme of the present invention will be further described in detail in connection with appended drawings and embodiments.
The core theory of the present invention is that: first, the user packet queues to be transmitted are divided into the user packet queues with lost packet and the user packet queues without lost packet; for the user packet queues with lost packet, if a real time lost ratio of packet for the user excesses a predetermined lost ratio threshold of packet, terminates a connection to the user; if the real time lost ratio of packet for the user does not excess the predetermined lost ratio threshold of packet, schedules the user packet queues according to a volume of the lost ratio of packet; for the user packet queues without lost packet, schedules according to packet lengths, channel quality states, time delays and time delay jitters.
As shown in the flowchart of
In the above information, the channel quality state is the reciprocal of the maximum transmission rate possible for the wireless channel, and is referred as Ci,j, wherein, i represents one user index, j represents the scheduling period, and its value is an integer not less than one. The channel quality state Ci,j can also be represented in other ways, for example, represented as the reciprocal of S/N of channel or the path loss of transmission.
All the length of the packets to be transmitted is referred as Ii,j, wherein, i represents one user index, j represents the scheduling period, the value of j is an integer not less than one.
The maximum delay threshold of various services is referred as Wmax,m, m corresponds to the type of services with limitation of maximum delay threshold. The delay waiting time of each packet is referred as Wi,j, wherein, i represents one user index, j represents the scheduling period, the value of j is an integer not less than one.
The real time lost ratio of packet of various users is referred as PLi,j, wherein i represents one user index, j represents the scheduling period, the value of j is an integer not less than one. The real time lost ratio of packet PLi,j can be represented by the proportion of lost packets to all of the transmitted packets in a certain period of time, the concrete length of time depends on the statistic period of the whole system, which can usually be selected between 200 ms and 2 s, with 200 ms is preferable, because the shorter the time is, the faster the response of system is, but since the consumption of system resource will be increased, an overall consideration is necessary. The real time lost ratio threshold of packet of each user is referred as PLmax, and PLmax>0.
The packet time delay jitter is referred as Jitteri,j, wherein i represents one user index, j represents the scheduling period, the value of j is an integer not less than one. The time delay jitter Jitteri,j can be represented in recursion:
Jitteri,j=0, when j=1,
Jitteri,j=α|(Wi,j+li,jCi,j)−(Wi,j−1+li,j−1Ci,j−1)|+(1−α)Jitteri,j−1, when j∈[2,+∞),
Wherein α∈(0,1). The time delay jitter Jitteri,j can also be represented in other forms, for example, the recursion of time delay jitter can be defined as Jitteri,j=α|(Wi,j)−(Wi,j−1)|+(1−α)Jitteri,j−1. The maximum time delay jitter threshold of packet can be represented as Jittermax,m, wherein n corresponds to the type of services with limitation of maximum time delay jitter threshold.
After obtaining the above related information, judge whether there are users whose real time lost ratio of packet PLi,j is more than 0 (step 102), if yes, judge whether there are users whose real time lost ratio of packet PLi,j excesses the real time lost ratio threshold of packet PLmax (step 103), for whom, the connection will be terminated, and step 107 will be performed to judge whether the packet queue is vacant; if there is no user whose real time lost ratio of packet PLi,j excesses the real time lost ratio threshold of packet PLmax, that is, when PLmax≧PLi,j>0, schedule the user packet according to the volume of lost ratio of user packet, that is, give priority to scheduling the user packet with big packet lost ratio, till all the user packets less than the real time lost ratio threshold of packet PLmax are scheduled, then perform step 107 to judge whether the packet queue is vacant.
If there is no user whose real time packet lost ratio PLi,j is more than 0, perform step 106, that is consider the packet length li,j, channel state Ci,j time delay Wi,j, time delay jitter Jitteri,j etc. comprehensively to schedule the packets in the queue, scheduling with priority can be performed according to the principal of least (Wmax,m−Wi,j)(Jittermax,m−Jitteri,j)li,jCi,j, wherein (Wmax,m−Wi,j) represents the limitation to time delay, (Jittermax,m−Jitteri,j) represents the limitation to the time delay jitter, li,j represents the consideration over the packet length, Ci,j represents the consideration over the channel state; The principal of least (Jittermax,n−Jitteri,j)li,jCi,j/Wi,j can also be employed for scheduling. After finishing the scheduling, judge whether the packet queue is vacant (step 107).
If the packet queue is vacant, end the scheduling (step 108); if not, turn to step 101 to perform the scheduling of next period, till all the packet queues to be transmitted are vacant.
The scheduling unit 203 buffers the data packet temporarily, calculates the delay waiting time Wi,j of each packet, the real time packet lost ratio PLi,j of each user, the time delay jitter of packet Jitteri,j, and receives the control information S211 transmitted from MAC controller 201, including the QoS requirements of services, the delay threshold Wmax,m the time delay jitter threshold Jittermax,n, the real time packet lost ratio threshold PLmax of each user, the delay occurred before MAC-hs and the overall power limit of all the HS-DSCH. The uplink signaling S212 from High Speed Downlink Shared Channel (simplified as HS-DSCH) 206 includes such channel state information as the maximum number of transmission bits in the transmission time interval TTI, the modulating manners, the code channel number, etc.. Then the packet date is scheduled in the scheduling unit 203 according to the scheduling method of the present invention.
After being scheduled, the packet data is transmitted to the HARQ module 204 for performing the hybrid feedback retransmission function, then to the TFRC module 205 for performing the transmission format selection function, and finally is assigned to HS-DSCH 206 to be transmitted in air.
Lastly, it should be noted that the above embodiments are only to illustrate the technical scheme of the present invention, without any limitation. Although the present invention is described in detail with reference to the preferred embodiments, the ordinary person skilled in the art should understand that the scheme of the present invention can be modified or substituted, without departing from the spirit and scope of the technical schemes of the present invention, all of which should be covered within the following claims.
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/CN03/00966 | 11/14/2003 | WO | 5/12/2006 |