CN100512055C - Mixed service resource distributing method for OFDM system - Google Patents

Abstract

The hybrid business resource allocation method in FDM system comprises: presenting a service quality satisfaction level criterion based on optimal physical and data linkage layers; with this criterion, according to active user number, server quality request and received service utility function value in every time slot, deciding user PRI, and allocating adaptively sub-carrier to user with high PRI. This invention can meet user service request and ensures fairness.

Description

Mixed service resource allocation method for orthogonal frequency division multiplexing system

The technical field is as follows:

the invention belongs to the technical field of Orthogonal Frequency Division Multiplexing (OFDM) mobile communication, and particularly relates to a hybrid service resource allocation method in an OFDM communication system.

Background art:

conventional OFDM resource allocation research focuses on both bit allocation and power allocation, such as the multi-user adaptive subcarrier, bit and power allocation algorithm mentioned in the international institute of electrical and electronics engineers (IEEE j.on Select areas. commun, Volume 17, No10, 1999, pp 1747-. Since these algorithms do not consider the random characteristics of the traffic arrival, the queue status of the user, the Quality of Service (QoS) requirements of the user, and the like in the allocation process, it is difficult to provide satisfactory Service to the user.

The method for distributing mixed resources, which is mentioned in the International Conference on institute of electrical and electronics engineers (IEEE communications society IEEE International Conference, Volume 1, 2004, pp58-62), does not provide a scheduling algorithm applicable to both real-time services and non-real-time services because the utility functions of the real-time services and the non-real-time services are respectively calculated by using an exponential scheduling criterion and a proportional fairness criterion; furthermore, this approach is also unreasonable to assume that data traffic packets are not dropped due to latency.

The invention content is as follows:

the invention provides a resource allocation method of an orthogonal frequency division multiplexing system, which considers the random characteristic of service achievement, the requirement of service on waiting time delay, the state of a user queue, the requirement of the user on service quality and the limitation of the system on the length of the user queue, is suitable for both real-time service and non-real-time service, can better meet the requirement of the user on service quality, and can ensure the fairness of the satisfaction degree of the service quality among users.

The invention relates to a mixed service resource allocation method of an orthogonal frequency division multiplexing system, wherein the orthogonal frequency division multiplexing system comprises a scheduling module and a self-adaptive subcarrier and bit combined allocation module; the base station provides a queue with a limited length for each user, wherein the queue length is the product of the peak arrival rate of a packet and the maximum tolerable latency, the base station also establishes and maintains a label for a service queue of each user, and the label records a corresponding user identifier, a packet identifier, a subcarrier identifier allocated to the user, the number of packets arrived in the user queue cumulatively, the number of packets lost cumulatively by the user, the number of packets received in error cumulatively by the user, the number of packets in the current user queue, the latency of a packet at the head of the queue, the maximum tolerable latency of the packet, the maximum tolerable packet loss rate of the packet, the maximum tolerable packet error rate of the packet and the maximum tolerable packet loss rate of the packet; the base station inputs the data packet of the service into an independent queue of each user, and the service sequence of the packet in the queue is first-come first-served; the base station end informs users with nonzero packet number in the queue to carry out channel estimation through a broadcast channel; the user reports the channel estimation result and the number of error packets received in the last resource allocation process to the base station through an error-free feedback channel; the resource allocation algorithm at the base station side is executed once in each time slot, only one subcarrier is allocated to one user in each allocation process, and each subcarrier equally divides the total transmitting power of the system; at the starting time of each time slot, the scheduling module sequences the users and provides the sequenced user identifier sequences and the packets in the user queues to the adaptive subcarrier and bit joint distribution module; the adaptive subcarrier and bit joint distribution module firstly carries out descending sequencing on the subcarriers of the users according to the gain according to the channel state information fed back by the users to obtain a subcarrier gain descending sequence of each user, distributes the subcarriers for the packets in the user queue according to the user identifier sequence provided by the scheduling module and the packet information in the user queue, and feeds back the number of the packets which can be transmitted by each user to the scheduling module;

the method is characterized in that:

the scheduling module generates a user utility functionA decreasing number sequence comprising: the initialization unit discards packets with the waiting time delay more than or equal to the maximum waiting time delay which can be endured by the packets according to the feedback of the adaptive subcarrier and bit combined distribution module after the last resource distribution is finished at the starting time of each time slot, counts the number of error packets which are accumulatively received by each user according to the number of error packets which are fed back by the user and are received in the last resource distribution process, clears the subcarrier identifier which is distributed by the user, establishes a new packet identifier for the packet which arrives in the last time slot, counts the number of the packets which are accumulatively arrived in each user queue, counting the number of packets lost by each user in an accumulated way, counting the number of packets left in a user queue, counting the total number of packets left in the user queue of the time slot, recording the waiting time delay of the head packet of each user queue and counting the total number of users in an activated state in the time slot system; after initialization, the satisfaction factor calculation unit follows the formula $S_k\left[i\right]={\mathrm{PLR}}_k\left[i\right]/{\mathrm{PLR}}_k^{\max }$ Calculating a satisfaction factor for each user k, wherein

Represents the maximum packet loss rate, PLR, that user k can tolerate_k[i]Indicating the packet loss rate, PLR, of user k at the start of the ith slot_k[i]The equation can be used

Making an estimate, wherein PLN_k[i]、PEN_k[i]And PAN_k[i]Respectively representing the accumulated lost packet number, the received error packet number and the total number of the arrived packets of the user k before the ith time slot; the utility function calculating unit calculates the satisfaction factor according to the formula <math> <mrow> <msubsup> <mi>U</mi> <mi>k</mi> <mi>CL</mi> </msubsup> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> <mo>=</mo> <msub> <mi>u</mi> <mi>k</mi> </msub> <mfrac> <mrow> <msubsup> <mi>R</mi> <mi>k</mi> <mi>cur</mi> </msubsup> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> <mrow> <msubsup> <mi>R</mi> <mi>k</mi> <mi>aver</mi> </msubsup> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> </mfrac> <mi>exp</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>S</mi> <mi>k</mi> </msub> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> <mo>-</mo> <mover> <mrow> <mi>S</mi> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> <mo>&OverBar;</mo> </mover> </mrow> <mrow> <mn>1</mn> <mo>+</mo> <msqrt> <mover> <mrow> <mi>S</mi> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> <mo>&OverBar;</mo> </mover> </msqrt> </mrow> </mfrac> <mo>)</mo> </mrow> </mrow></math> Calculating a utility function for a non-zero number of users k per queue, wherein

Indicating the highest data transmission rate that user k predicts can support in the ith slot,

indicating that user k is at the ith timeAverage data transfer rate before slot, using formula $R_k^{\mathrm{aver}}[i+1]=(1-\frac{1}{t_c})R_k^{\mathrm{aver}}\left[i\right]+\frac{1}{t_c}{R}_k^{\mathrm{cur}}\left[i\right]$ Performing an update, wherein t_cIndicating the length of the sliding time window, if the base station does not transmit data to user k in the ith time slotUse of $R_k^{\mathrm{cur}}\left[i\right]=0$ Updating is carried out; u. of_kUse of $u_k=-\log \left({\mathrm{PDR}}_k^{\max }\right)/D_k^{\max }$ Performing a calculation in which

And

respectively representing the maximum packet loss rate and the packet delay tolerable by the user k, Si]The mean value of the satisfaction degree of the activated users in the system before the ith time slot is represented by the formula <math> <mrow> <mover> <mrow> <mi>S</mi> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> <mo>&OverBar;</mo> </mover> <mo>=</mo> <mfrac> <mn>1</mn> <mi>K</mi> </mfrac> <munderover> <mi>&Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>K</mi> </munderover> <msub> <mi>S</mi> <mi>k</mi> </msub> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> <mo>,</mo> </mrow></math> Wherein K represents the total number of users in the active state in the ith time slot system; after the calculation is finished, the sequencing unit performs descending sequencing on the utility function of the user to obtain a descending sequence of the utility function of the user, and then the joint allocation process of the self-adaptive subcarrier and the bit is performed; the adaptive subcarrier and bit joint distribution module comprises: the subcarrier sequencing unit is used for sequencing the subcarriers of the users with nonzero packet quantity in a descending manner according to the gain according to the channel state information fed back by the users to obtain a subcarrier gain descending sequence of each user; the sub-carrier and power combined allocation unit allocates idle sub-carriers of the user sub-carrier gain decreasing sequence for the packets in the user queue in sequence from the head user in the user utility function decreasing sequence, records the identifier of the sub-carrier allocated to each user, calculates and records the transmitting bit number allocated to the sub-carriers until no packet of the user needs to be transmitted or the number of idle sub-carriers of the system is zero; if the system has idle sub-carriers and has packets to be transmitted, continuously distributing the idle sub-carriers in the user sub-carrier gain decreasing sequence for the packets in the next user queue with lower priority in the user utility function decreasing sequence until no packets need to be transmitted or the number of the idle sub-carriers in the system is zero; dispensing knotAfter the resource is bound, the statistic feedback unit records the result of the current resource allocation and feeds back the number of the packets which can be transmitted by each user to the initialization unit of the scheduling module.

The working principle of the scheduling module is as follows:

since different services have different requirements on delay, packet loss rate and packet error rate, the maximum packet loss rate that they can tolerate may be different. From the perspective of fairness, for different types of services, the service satisfaction degree of the service to the service which has been received is embodied by using the factor obtained by normalizing the service quality which has been received by the service by the worst service quality which can be tolerated by the service, which is called as a service quality satisfaction factor in the invention. And a fairness criterion based on the factor is referred to as a quality of service satisfaction fairness criterion.

The quality of service satisfaction factor may be expressed as:

$S_k\left[i\right]={\mathrm{PLR}}_k\left[i\right]/{\mathrm{PLR}}_k^{\max }---\left(1\right)$

wherein, PLR_k[i]Representing the packet loss rate counted by the user k at the start time of the ith slot, which is calculated as shown in equation (2);

represents the maximum packet loss rate that user k can tolerate, which is calculated as shown in equation (3);

${\mathrm{PLR}}_k\left[i\right]=\frac{{\mathrm{PLN}}_k\left[i\right]+{\mathrm{PEN}}_k\left[i\right]}{{\mathrm{PAN}}_k\left[i\right]}---\left(2\right)$

wherein, PLN_k[i]、PEN_k[i]And PAN_k[i]Respectively representing the accumulated lost packet number, the received error packet number and the arrived packet number of the user k before the ith time slot.

${\mathrm{PLR}}_k^{\max }=1-(1-{\mathrm{PER}}_k^{\max })(1-{\mathrm{PDR}}_k^{\max })---\left(3\right)$

Wherein,and

respectively, the packet error rate caused by the wireless channel and the packet loss rate caused by too large packet delay, which can be tolerated by the user k.

The quality of service satisfaction factor is applicable to both real-time and non-real-time services. On one hand, the method embodies the satisfaction degree of the users to the service provided by the system, and on the other hand, embodies the fairness of the service quality among the users. The smaller the service quality satisfaction factor is, the more satisfied the user is with the service provided by the system; conversely, the less satisfactory the user is with respect to the services provided by the system. The smaller the difference of the service quality satisfaction degree factors among a plurality of users is, the more fair the service quality provided by the system to different users is reflected; conversely, the less fair the quality of service provided by the system to different users. In a user QoS centric system, the use of the quality of service satisfaction fairness criterion is more reasonable than the use of user latency or user average throughput as the criterion for system fairness. Because the user waiting time delay is used as the criterion of the system fairness, the user waiting time delay can be regarded as the fairness criterion for neglecting the influence of error transmission brought by a wireless channel on the user performance; using the average throughput of users as a criterion for system fairness can be regarded as a fairness criterion assuming that user packets have no latency constraints and user queues are infinitely long. They are all special cases of quality of service fairness criteria.

Based on the fairness criterion of the satisfaction degree of the service quality, the resource allocation method of the invention defines the utility function of the user as

<math> <mrow> <msubsup> <mi>U</mi> <mi>k</mi> <mi>CL</mi> </msubsup> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> <mover> <mo>=</mo> <mi>&Delta;</mi> </mover> <msub> <mi>u</mi> <mi>k</mi> </msub> <mfrac> <mrow> <msubsup> <mi>R</mi> <mi>k</mi> <mi>cur</mi> </msubsup> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> <mrow> <msubsup> <mi>R</mi> <mi>k</mi> <mi>aver</mi> </msubsup> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> </mfrac> <mi>exp</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>S</mi> <mi>k</mi> </msub> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> <mo>-</mo> <mover> <mrow> <mi>S</mi> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> <mo>&OverBar;</mo> </mover> </mrow> <mrow> <mn>1</mn> <mo>+</mo> <msqrt> <mover> <mrow> <mi>S</mi> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> <mo>&OverBar;</mo> </mover> </msqrt> </mrow> </mfrac> <mo>)</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </mrow></math>

Wherein,

represents the highest data transmission rate that user k predicts can support in the ith slot; u. of_kIs a weighting coefficient, which is calculated as shown in equation (5); s [ i ]]The average value of the satisfaction degree of the activated users in the system before the ith time slot is shown in the formula (6);

represents the average data transmission rate of user k before the ith slot, and is calculated as shown in formula (7);

$u_k=-\log \left({\mathrm{PDR}}_k^{\max }\right)/D_k^{\max }---\left(5\right)$

wherein,

represents the maximum packet delay that user k can tolerate;

<math> <mrow> <mover> <mrow> <mi>S</mi> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> </mrow> <mo>&OverBar;</mo> </mover> <mo>=</mo> <mfrac> <mn>1</mn> <mi>K</mi> </mfrac> <munderover> <mi>&Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>K</mi> </munderover> <msub> <mi>S</mi> <mi>k</mi> </msub> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>6</mn> <mo>)</mo> </mrow> </mrow></math>

wherein, K represents the total number of users in the activation state in the ith time slot system;

$R_k^{\mathrm{aver}}[i+1]=(1-\frac{1}{t_c})R_k^{\mathrm{aver}}\left[i\right]+\frac{1}{t_c}{R}_k^{\mathrm{cur}}\left[i\right]---\left(7\right)$

wherein, t_cRepresents the length of the sliding time window; if the base station does not transmit data to user k in the ith time slot $R_k^{\mathrm{cur}}\left[i\right]=0.$

t_cThe size of the data reflects the tolerance degree of the user to the data which cannot be received for a long time; the longer the window, the longer the waiting time the user can tolerate; conversely, the shorter the time that can be tolerated.

The scheduling algorithm only calculates the utility function of the user at the beginning of each time slot, and the data of the user can arrive at any time in one time slot, so that the packet arriving in the time slot only influences the utility function calculated in the next time slot of the user. Because the invention considers the random character of the packet arrival of the user of the data link layer, if a certain user does not receive the data transmitted by the base station in the ith time slot, the following three reasons can be caused:

no data packet in the user queue waits for transmission;

the number of packets to be transmitted in the user queue is not zero, but the user does not get the opportunity to select the sub-carrier;

the data in the user queue is not enough to constitute a packet; for example, when voice traffic is active, data arrives at a fixed rate of 64kbps, each packet consists of 1280bits, and the scheduler updates once every millisecond, then the arrival of 64 bits of voice traffic in 1ms is insufficient to constitute a packet, so that the user may not have packets to be transmitted for several scheduling intervals.

The first user selected by the resource allocation method of the invention is

$k=\underset{j=1,L,K}{\arg \max }{U}_j^{\mathrm{CL}}\left[i\right]---\left(8\right)$

As can be seen from the formula (4), the formula (7) and the formula (8), if the waiting delay of a user is large, the packet loss rate and the packet loss rate are both high, so that the service quality satisfaction factor of the user is also high, and the utility function of the user is improved; if the channel quality of a user is poor, the packet loss rate and the packet error rate are also high, and the utility function of the user can be improved.

In this way, the scheduling module may obtain a decreasing sequence of user utility functions

Wherein <math> <mrow> <mo>&ForAll;</mo> <mi>k</mi> <mo>&prime;</mo> <mo>&lt;</mo> <mi>k</mi> <mo>,</mo> </mrow></math> k′，k∈[1，K]Existence of U_k′[i]>U_k[i]。

The adaptive subcarrier and bit joint allocation algorithm adaptively allocates subcarriers and the number of bits carried on each subcarrier to the user according to the channel state information fed back by the user, the user sequencing information provided by the scheduling algorithm and the number of remaining packets in the user queue. The goal is to optimize the system throughput using adaptive modulation techniques given the transmit power per subcarrier while ensuring that the packet error rate counted by each user does not exceed the maximum packet error rate that the user can tolerate. The working principle is as follows:

the adaptive subcarrier and bit joint distribution module firstly decreases the sequence for the user utility function

The first user in the queue allocates the sub-carriers, when the user has no remaining packets to be transmitted, the sub-carriers are continuously allocated to the next user with lower priority until no packets need to be transmitted or the system has no idle sub-carriers. The method has the advantage that the satisfaction degree of the user who is not satisfied with the service provided by the system at present can be improved.

The received signal-to-noise ratio at the receiving end can be described using the following equation:

<math> <mrow> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>b</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>n</mi> </mrow> </msub> <mo>)</mo> </mrow> <mo>=</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>K</mi> </munderover> <mfrac> <mrow> <msub> <mi>P</mi> <mi>total</mi> </msub> <mo>&times;</mo> <msup> <mrow> <mo>|</mo> <msubsup> <mi>h</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>n</mi> </mrow> <mi>k</mi> </msubsup> <mo>|</mo> </mrow> <mn>2</mn> </msup> </mrow> <mrow> <mi>N</mi> <mo>&times;</mo> <msubsup> <mi>&sigma;</mi> <mi>n</mi> <mn>2</mn> </msubsup> </mrow> </mfrac> <mo>&times;</mo> <msubsup> <mi>A</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>n</mi> </mrow> <mi>k</mi> </msubsup> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>9</mn> <mo>)</mo> </mrow> </mrow></math>

wherein, P_totalIs the total power transmitted by the base station; $A_{i,n}^k=1$ indicating that the user k is allocated to the nth subcarrier at the ith time slot; because the receiving party can not detect from the same subcarrier at the same timeData of a plurality of users, so for an arbitrary subcarrier n $A_{i,n}^k=1$ When there is <math> <mrow> <msubsup> <mi>A</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>n</mi> </mrow> <mrow> <mi>k</mi> <mo>&prime;</mo> </mrow> </msubsup> <mo>=</mo> <mn>0</mn> </mrow></math> <math> <mrow> <mo>&ForAll;</mo> <mi>k</mi> <mo>&prime;</mo> <mo>&NotEqual;</mo> <mi>k k</mi> <mo>,</mo> <mi>k</mi> <mo>&prime;</mo> <mo>&Element;</mo> <mrow> <mo>[</mo> <mn>1</mn> <mo>,</mo> <mi>K</mi> <mo>]</mo> </mrow> <mo>;</mo> </mrow></math>

Representing the amplitude of the channel gain on the nth subcarrier estimated by the user k at the starting moment of the ith time slot;for the noise power on each subcarrier, b_i,nThe number of information bits carried by the nth subcarrier in the ith slot is represented, and the calculation is shown in formula (10):

<math> <mrow> <msub> <mi>b</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>n</mi> </mrow> </msub> <mo>=</mo> <mi>&Delta;</mi> <msub> <mrow> <mi>f</mi> <mi>log</mi> </mrow> <mn>2</mn> </msub> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mfrac> <mrow> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>b</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>n</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> <mi>&Gamma;</mi> </mfrac> <mo>)</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>10</mn> <mo>)</mo> </mrow> </mrow></math>

wherein for MQAM modulation, Γ may be used

And (4) approximation.

Compared with the prior art, the invention considers the random characteristic of service achievement, the requirement of service on waiting time delay, the requirement of service on packet loss rate, the requirement of service on packet error rate, the user queue state and the limitation of a system on the user queue length in the resource allocation process, and provides a service quality satisfaction fairness criterion by using a cross-layer criterion, wherein the criterion is suitable for both real-time service and non-real-time service, and is a cross-layer fairness criterion based on the joint optimization of a physical layer and a data link layer. According to the analysis of the criterion, the resource allocation method provided by the invention uses the utility function suitable for the mixed service to represent the priority of the user, i.e. the user with unsatisfactory service provided by the system is endowed with higher priority, and the adaptive subcarrier and bit joint allocation module preferentially allocates subcarriers to the user with high level. Therefore, the invention not only can better meet the requirement of users on the service quality, but also can ensure the fairness of the satisfaction degree of the service quality among the users.

Description of the drawings:

fig. 1 is a schematic block diagram of a resource allocation method of an ofdm system according to the present invention.

FIG. 2 is a simulated statistical histogram of packet loss rate for a streaming media user when the number of users is 18;

fig. 3 is a simulated statistical histogram of packet loss rate of streaming media users when the number of users is 22.

The specific implementation mode is as follows:

embodiments of the method are described below with reference to the drawings.

Example 1:

the present embodiment uses an OFDM transmission system having a bandwidth of 2.048MHz and a number of data subcarriers of 64. The channel adopts a COST207 six-path model, the maximum multi-path time delay is 10us, the six paths are distributed at equal intervals within 0-10 us, and the power spectral density of each path meets the common Jake model. The maximum doppler frequency offset is 300 Hz. The time slot length is 1ms, the packet size is fixed to 1280bits, and the length t of the sliding window_cIs 1000 time slots. And adopting an MQAM adaptive modulation mode, wherein M is {0, 2, 4, 8, 16, 32 }. The number of voice users and the number of data users in the fixed system are 5 and 3, respectively.

The service source model adopts an ON-OFF two-state Markov process. The active and quiet periods are independent of each other and follow an exponential distribution. The transition probability from the active period to the silent period is γ ═ 1-exp (-T)_slot/t₁) The transition probability from the silent period to the active period is u-1-exp (-T)_slot/t₂) Wherein T is_slotIs the time slot length, t₁And t₂The mean of the active and quiet periods, respectively. The business model parameters used are shown in table 1 below:

TABLE 1 Business model parameters

System parameter	Numerical value	System parameter	Numerical value
				Mean burst length of speech	1.0s	Length of voice average silence	1.35s
Maximum PDR allowed for voice	10<sup>-2</sup>	Maximum PER allowed for voice	10<sup>-2</sup>
				Maximum delay allowed for voice	50ms	Average rate of voice traffic	64Kbps
Average burst length of streaming media	5.0s	Average silence length of streaming media	4.05s
				Maximum PDR allowed for streaming media	10<sup>-3</sup>	Maximum PER allowed for streaming media	10<sup>-3</sup>
Maximum delay allowed for streaming media	600ms	Average rate of streaming media service	640Kbps
				Data average burst length	10.0s	Data average silence length	8.0s
Maximum PDR allowed for data	10<sup>-3</sup>	Maximum PER allowed for data	10<sup>-3</sup>
				Maximum delay allowed for data	2s	Average rate of data traffic	1.28Mbps

Fig. 1 shows a schematic block diagram of a resource allocation method of an ofdm system according to the present invention: at the base station end, at the beginning of each time slotThe initialization unit 1 updates the label of each access user queue, which comprises discarding the packet with the waiting time delay larger than or equal to the maximum waiting time delay that can be endured by the packet according to the packet number signal 17 that can be transmitted by each user and fed back by the statistical feedback unit 16 after the last resource allocation is finished, counting the number of the error packets that are cumulatively received by each user according to the number of the error packets that are received in the last resource allocation process and fed back by the user, clearing the sub-carrier identifier that is allocated by the user, establishing a new packet identifier for the packet that arrives in the last time slot, counting the number of the packets cumulatively arrived in each user queue, counting the number of the packets cumulatively lost by each user, counting the number of the remaining packets in the user queue, counting the total number of the remaining packets in the user queue in the time slot, recording the waiting time delay of the head packet of each user queue and counting the total number of, sending an initialization ending signal 2 after the statistics is ended; the user subcarrier ordering unit 8 starts to order the subcarriers of the users in a decreasing manner according to the initialization ending signal 2 sent by the initialization module to obtain a subcarrier gain decreasing sequence of each user; the satisfaction factor calculating unit 3 starts to calculate the satisfaction factor of each user according to the formula (1), the formula (2) and the formula (3) according to the initialization ending signal 2 sent by the initialization module; the utility function calculation unit 4 calculates the utility function of the users with nonzero packet number in the queue according to the satisfaction factor of each user provided by the satisfaction factor calculation unit and according to the formula (4), the formula (5), the formula (6) and the formula (7); after the calculation is finished, the utility function sorting unit 5 sorts the utility functions of the users in a descending manner to obtain a user utility function descending sequenceAnd will be

I.e. the signal 6 is sent to the subcarrier and bit joint allocation unit 7; the sub-carrier and bit combined allocation module 7 allocates the sub-carrier of the user to the group in the user queue in turn from the head user of the queue according to the fed user utility function decreasing sequence signal 6 and the sub-carrier gain decreasing sequence of each user stored in the user sub-carrier ordering module 8Idle subcarriers in the gain decreasing sequence, record the identifier of the subcarrier allocated to each user, calculate and store the bit number of information carried by the subcarriers according to formula (9) and formula (10), record the total number of packets that can be transmitted in the time slot until the user has no packets to be transmitted or the number of idle subcarriers in the system is zero, send the number of the allocated subcarriers, i.e., signal 10, to the subcarrier number decision unit 11, and send the total number of packets that can be transmitted in the time slot, i.e., signal 9, to the system transmission packet number decision unit 14; the sub-carrier number judging module 11 judges whether the received judgment value is equal to the total number of the system sub-carriers, if not, the triggering signal 12 is sent to the system transmission grouping number judging unit 14; if the two signals are equal, generating a sub-carrier and bit joint distribution end signal 13 and sending the sub-carrier and bit joint distribution end signal into a statistical feedback unit 16; triggering a system transmission packet quantity judgment module 14 by a trigger signal 12 to judge whether a received judgment value 9 is equal to the total number of packets left in the current time slot of the system, if not, generating a trigger signal 15 and sending the trigger signal to a subcarrier and bit joint distribution module 7, wherein the subcarrier and bit joint distribution module 7 continuously distributes idle subcarriers in a user subcarrier gain decreasing sequence for packets in a next user queue with lower priority in the user utility function decreasing sequence according to the signal, recording identifiers of the subcarriers distributed by each user, calculating and storing information bit numbers carried by the subcarriers according to a formula (9) and a formula (10), and sending out a subcarrier and bit joint distribution end signal 13 until all users in the system do not need to be transmitted or the number of idle subcarriers in the system is zero; the subcarrier and bit joint allocation end signal 13 triggers the statistical feedback unit 16 to record the identifier of the subcarrier allocated to each user and the number of information bits allocated to each subcarrier, and feeds back the number of packets that each user can transmit, i.e. the signal 17, to the initialization module 1.

When 10 streaming media users exist in the system, the algorithm provided by the invention can not only maintain the packet loss rate of the users at a lower level, but also ensure the fairness of the satisfaction degree of the service quality among the users; when the number of streaming media users increases to 14Since the service satisfaction factor of the user is the exponential weighting factor of the utility function, the variation of the user packet loss rate is small and maintained at 5 × 10^-4～7×10^-4Such a lower level has better performance. The following features are shown from the simulated statistical histograms of packet loss rates for streaming media users with the number of users 18 and 22 given in fig. 2 and 3, respectively:

1. when the number of different users is different, the fairness of the satisfaction degree of the service quality among the users can be ensured.

2. The user packet loss rate fluctuates little with the change of the number of users.

3. As the number of users increases, the user packet loss rate remains at a lower level.

The invention takes the improvement of the satisfaction degree of a user to a system as a starting point, provides a cross-layer resource allocation algorithm for the downlink mixed service of an OFDM system, and considers the random characteristic of service achievement, the requirement of the service on waiting time delay, the requirement of the service on packet loss rate, the requirement of the service on packet error rate, the user queue state and the limitation of the system on the user queue length. By using a cross-layer criterion, the invention provides a service quality satisfaction fairness criterion, considers the combined influence of a data link layer and a physical layer on the service quality received by a user, and is suitable for both real-time services and non-real-time services; on the basis, the resource allocation algorithm provided by the invention uses the utility function to represent the priority of the user, i.e. the user with unsatisfactory service provided by the system is endowed with higher priority, and the adaptive subcarrier and bit joint allocation module preferentially allocates subcarriers to the user with high grade. Therefore, the invention not only can better meet the requirement of users on the service quality, but also can ensure the fairness of the satisfaction degree of the service quality among the users.

Claims (1)

1. A mixed service resource allocation method of an orthogonal frequency division multiplexing system comprises a scheduling module and a self-adaptive subcarrier and bit combined allocation module; the base station provides a queue with a limited length for each user, wherein the queue length is the product of the peak arrival rate of a packet and the maximum tolerable latency, the base station also establishes and maintains a label for a service queue of each user, and the label records a corresponding user identifier, a packet identifier, a subcarrier identifier allocated to the user, the number of packets arrived in the user queue cumulatively, the number of packets lost cumulatively by the user, the number of packets received in error cumulatively by the user, the number of packets in the current user queue, the latency of a packet at the head of the queue, the maximum tolerable latency of the packet, the maximum tolerable packet loss rate of the packet, the maximum tolerable packet error rate of the packet and the maximum tolerable packet loss rate of the packet; the base station inputs the data packet of the service into an independent queue of each user, and the service sequence of the packet in the queue is first-come first-served; the base station end informs users with nonzero packet number in the queue to carry out channel estimation through a broadcast channel; the user reports the channel estimation result and the number of error packets received in the last resource allocation process to the base station through an error-free feedback channel; the resource allocation algorithm at the base station side is executed once in each time slot, only one subcarrier is allocated to one user in each allocation process, and each subcarrier equally divides the total transmitting power of the system; at the starting time of each time slot, the scheduling module sequences the users and provides the sequenced user identifier sequences and the packets in the user queues to the adaptive subcarrier and bit joint distribution module; the adaptive subcarrier and bit joint distribution module firstly carries out descending sequencing on the subcarriers of the users according to the gain according to the channel state information fed back by the users to obtain a subcarrier gain descending sequence of each user, distributes the subcarriers for the packets in the user queue according to the user identifier sequence provided by the scheduling module and the packet information in the user queue, and feeds back the number of the packets which can be transmitted by each user to the scheduling module;

the method is characterized in that:

the scheduling module generates a decreasing sequence of user utility functions, comprising: the initialization unit discards a packet with the waiting time delay more than or equal to the maximum waiting time delay which can be endured by the packet according to the feedback of the self-adaptive subcarrier and bit combined distribution module after the last resource distribution is finished at the starting time of each time slot, counts the number of error packets which are received by each user in an accumulated mode according to the number of error packets which are received by the user in the last resource distribution process, clears the subcarrier identifier distributed by the user, establishes a new packet identifier for the packet which arrives in the last time slot, counts the number of the packets which arrive in each user queue in an accumulated mode, and counts the number of the packets which arrive in each user queue in an accumulated modeThe method comprises the steps that each user accumulates lost packet quantity, counts the residual packet quantity of a user queue, counts the residual packet quantity of the user queue of the time slot, records the waiting time delay of the head packet of each user queue and counts the total number of the users in an activated state in the time slot system; after initialization, the satisfaction factor calculation unit follows the formula $S_k\left[i\right]={\mathrm{PLR}}_k\left[i\right]/{\mathrm{PLR}}_k^{\max }$ Calculating a satisfaction factor for each user k, wherein

Represents the maximum packet loss rate, PLR, that user k can tolerate_k[i]Indicating the packet loss rate, PLR, of user k at the start of the ith slot_k[i]The equation can be usedMaking an estimate, wherein PLN_k[i]、PEN_k[i]And PAN_k[i]Respectively representing the accumulated lost packet number, the received error packet number and the total number of the arrived packets of the user k before the ith time slot; the utility function calculating unit calculates the satisfaction factor according to the formula <math> <mrow> <msubsup> <mi>U</mi> <mi>k</mi> <mi>CL</mi> </msubsup> <mrow> <mo></mo> <mo>[</mo> <mi>i</mi> <mo>]</mo> <mo></mo> </mrow> <mo>=</mo> <msub> <mi>u</mi> <mi>k</mi> </msub> <mfrac> <mrow> <msubsup> <mi>R</mi> <mi>k</mi> <mi>cur</mi> </msubsup> <mrow> <mo></mo> <mo>[</mo> <mi>i</mi> <mo>]</mo> <mo></mo> </mrow> </mrow> <mrow> <msubsup> <mi>R</mi> <mi>k</mi> <mi>aver</mi> </msubsup> <mrow> <mo></mo> <mo>[</mo> <mi>i</mi> <mo>]</mo> <mo></mo> </mrow> </mrow> </mfrac> <mi>exp</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>S</mi> <mi>k</mi> </msub> <mrow> <mo></mo> <mo>[</mo> <mi>i</mi> <mo>]</mo> <mo></mo> </mrow> <mo>-</mo> <mover> <mrow> <mi>S</mi> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> <mo>&OverBar;</mo> </mover> </mrow> <mrow> <mn>1</mn> <mo>+</mo> <msqrt> <mover> <mrow> <mi>S</mi> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> <mo>&OverBar;</mo> </mover> </msqrt> </mrow> </mfrac> <mo>)</mo> </mrow> </mrow></math> Calculating a utility function for a non-zero number of users k per queue, wherein

Indicating the highest data transmission rate that user k predicts can support in the ith slot,

representing the average data transmission rate of user k before the ith slot using the formula $R_k^{\mathrm{aver}}[i+1]=(1-\frac{1}{t_c})R_k^{\mathrm{aver}}\left[i\right]+\frac{1}{t_c}{R}_k^{\mathrm{cur}}\left[i\right]$ Performing an update, wherein t_cIndicating the length of the sliding time window, if the base station does not transmit data to user k in the ith time slot

Use of $R_k^{\mathrm{cur}}\left[i\right]=0$ Updating is carried out; u. of_kUse of $u_k=-\log \left({\mathrm{PDR}}_k^{\max }\right)/D_k^{\max }$ Performing a calculation in whichRespectively representing the maximum packet loss rate and the packet delay tolerable by the user k, Si]The mean value of the satisfaction degree of the activated users in the system before the ith time slot is represented by the formula <math> <mrow> <mover> <mrow> <mi>S</mi> <mrow> <mo></mo> <mo>[</mo> <mi>i</mi> <mo>]</mo> <mo></mo> </mrow> </mrow> <mo>&OverBar;</mo> </mover> <mo>=</mo> <mfrac> <mn>1</mn> <mi>K</mi> </mfrac> <munderover> <mi>&Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>K</mi> </munderover> <msub> <mi>S</mi> <mi>k</mi> </msub> <mrow> <mo>[</mo> <mi>i</mi> <mo>]</mo> </mrow> <mo>,</mo> </mrow></math> Wherein K represents the total number of users in the active state in the ith time slot system; after the calculation is finished, the sequencing unit performs descending sequencing on the utility function of the user to obtain a descending sequence of the utility function of the user, and then the joint allocation process of the self-adaptive subcarrier and the bit is performed; the adaptive subcarrier and bit joint distribution module comprises: the subcarrier sequencing unit is used for sequencing the subcarriers of the users with nonzero packet quantity in a descending manner according to the gain according to the channel state information fed back by the users to obtain a subcarrier gain descending sequence of each user; the sub-carrier and power combined allocation unit allocates idle sub-carriers of the user sub-carrier gain decreasing sequence for the packets in the user queue in sequence from the head user in the user utility function decreasing sequence, records the identifier of the sub-carrier allocated to each user, calculates and records the transmitting bit number allocated to the sub-carriers until no packet of the user needs to be transmitted or the number of idle sub-carriers of the system is zero; if the system has idle sub-carriers and has packets to be transmitted, continuously distributing the idle sub-carriers in the user sub-carrier gain decreasing sequence for the packets in the next user queue with lower priority in the user utility function decreasing sequence until no packets need to be transmitted or the number of the idle sub-carriers in the system is zero; after the distribution is finished, the statistic feedback unit records the result of the resource distribution and feeds back the number of the packets which can be transmitted by each user to the initialization unit of the scheduling module.

Images