CN106507463A - A resource allocation method for heterogeneous cellular networks based on multi-channel heuristic clustering - Google Patents

Abstract

The invention discloses a kind of isomery cellular network resource distribution method based on the heuristic sub-clustering of multichannel, in clustering process, the method allows one family base station to add in multiple clusters and carry out combined channel distribution with grand user, Home eNodeB is made to obtain multiple channels, then non-cooperative game model is set up for Home eNodeB transmission power on each channel in two-tier network, the utility function of definition considers not only the effectiveness of the domestic consumer, it is also contemplated that interference to domestic consumer and Hong user.Step is as follows：Oriented interference figure is set up for home base station network first, and Home eNodeB is added in multiple clusters according to the heuristic cluster algorithm of multichannel, distribute channel for grand user afterwards and match with cluster, shared channel；The game of the transmission power of each channel of Home eNodeB is then set up, optimal power set when obtaining balanced after successive ignition, Home eNodeB is according to set adjustment transmission power.

Description

A resource allocation method for heterogeneous cellular networks based on multi-channel heuristic clustering

technical field

The invention belongs to the technical field of communication, and relates to a heterogeneous cellular network resource allocation method based on multi-channel heuristic clustering.

Background technique

Spectrum resources in a wireless communication system are limited. In order to serve as many users as possible, a base station allocates the same spectrum resources to many users. This method can make full use of spectrum resources, but introduces interference between users using the same spectrum. When the user receives too much interference, it will cause errors in the transmission information and even cause communication interruption. Therefore, it is necessary to reduce the interference between users through reasonable resource allocation. In a hierarchical heterogeneous network, the home base station network and the macro cell simultaneously provide wireless communication services for users within the coverage area. In order to reduce co-channel interference, it is necessary to allocate and power control the channels used by each user.

In a hierarchical heterogeneous network, the home base station network and the macro cell provide wireless communication services for users within the coverage area at the same time. In order to reduce co-channel interference, it is necessary to allocate the channels used by each user. Among them, the heuristic analysis based on graph theory Cluster algorithm allocation channel is a relatively simple implementation. The algorithm divides the femtocells into multiple clusters and allocates channels jointly with macro users to achieve channel multiplexing, reducing the same-layer interference and cross-layer interference between users. However, this algorithm only allocates one channel for the home base station, which cannot meet the requirements of multiple users of the home base station, and the spectrum utilization rate is not high enough. At the same time, there are few researches on the power control on the Femtocell channel in the existing researches, most of which only consider the case of one home user per Femtocell.

Therefore, for the hierarchical heterogeneous network downlink scenario, the present invention improves the clustering algorithm based on graph theory, uses multi-channel heuristic clustering algorithm and game theory to analyze and manage interference in the network, and uses reasonable resource allocation To achieve the purpose of ensuring user performance and improving spectrum utilization. The present invention uses a multi-channel heuristic clustering algorithm to enable the home base station to join multiple clusters, and enables the home base station to use multiple channels through the joint channel allocation of macro users and home users, and then establishes a non-cooperative game model for each channel of the home base station The transmission power of the game is played, the interference caused by the home base station is controlled, and the communication performance of the user is guaranteed.

Contents of the invention

Technical problem: The purpose of this invention is to provide a resource allocation method based on multi-channel heuristic clustering that can make full use of spectrum resources in a layered heterogeneous network, reduce same-layer interference and cross-layer interference, and ensure the communication quality of all users .

Technical solution: The heterogeneous cellular network resource allocation method based on multi-channel heuristic clustering of the present invention includes the following steps:

1) Joint channel allocation

1.1) Define the home base station set {V}={v ₁ ,v ₂ ,...,v _F }, {v _i } represents home base station i, the number of home base stations is F, the number of macro users and the number of channels equal, both are M, the initial transmission power of the home base station is the same, the transmission power of each home base station is P _i , and the channels allocated by the home base station i are T _i ;

1.2) Home base station i calculates the interference weight of home base station j, the formula is as follows:

in is the channel gain of the interference signal of the home base station j to the user of the home base station i, The channel gain of the useful signal of the home base station i to the user of the base station, δ is the threshold of the weight, less than this threshold can be regarded as no interference between the two home base stations, through the formula Calculate the sum of the interference weights suffered by the home base station i to form a set of interference weights {W}={w ₁ ,w ₂ ,...,w _F };

1.3) Arranging the home base stations in descending order according to the interference weight w _i to form a new set {V'} of home base stations sorted;

1.4) Divide the home base stations into clusters {C _n } _n∈M according to the number of channels M:

1.4.1 Take a home base station i from the home base station set {V'} in turn to join cluster C _n , the number of home base stations that cluster C _n will interfere with home base station i is {v _i '} is the set of base stations that interfere with home base station i in the cluster C _n , and the sum of the added weights of the cluster C _n is Select the first T _i clusters with the smallest W _l ⁿ from all the clusters to add the home base station, and delete the home base station from the home base station set {V'};

1.4.2 Repeat step 1.4.1 until the home base station set {V'} is empty, then the home base station clustering is completed;

1.5) Joint channel allocation between home base stations and macro users:

1.5.1 The number of macro users is M, and the set of macro users is {Mu}. Arrange the macro users in the order from the closest to the farthest from the macro base station to obtain a new set of macro users {Mu'}, and allocate them to the macro users in sequence a channel;

1.5.2 Take a macro user m from the macro user set {Mu'} in order, and calculate the sum of the interference and noise it receives from the cluster C _n is the channel gain of the interference signal from the home base station i to the macro user m, select a cluster C _n' from the cluster set {C _n } that makes IN _n the smallest, match the macro user m with this cluster, the home base station in the cluster and the macro user m The user uses the same channel, and the home base station cluster and macro user are deleted from the home base station cluster set {C _n } and the macro user set {Mu'};

1.5.3 Repeat step 1.5.2 until both the home base station cluster set {C _n } and the macro user set {Mu'} are empty, and the joint channel allocation is completed;

2) Allocate the transmit power of each channel of the home base station

2.1) Initialization: set ε=10 ^-5 , time=0, the transmission power of each channel of the home base station is {P _i,t } ^(time) ={Pmax/T _i }, where P max is the maximum transmission power of the home base station, T _i is the number of channels used by home base station i;

2.2) Calculate the transmit power of each channel of the home base station, the formula is as follows:

where B is the channel bandwidth, k is the coefficient of the penalty term, is the channel gain of the interference signal of the home base station i to the macro user m, is the sum of the interference signal channel gains of home base station i to other home users using channel t, is the signal-to-interference-noise ratio of femtocell i on channel t, and the calculation formula is C _n is the cluster of femtocells using channel t, is the sum of the interference of other Femtocells in the cluster C _n to the user of Femtocell i on the channel t, P _m is the total transmit power of the macro base station, M is the number of macro users, is the channel gain of the interference signal from the macro base station to the user of the femtocell i, is the interference caused by the macro base station, N0 is the noise power spectral density of the system, B is the channel bandwidth, and N0*B is the noise on the channel. The transmission power on each channel is limited, 0<P _i,t ≤ Pmax/T _i , Pmax is the maximum transmission power of the home base station, and the maximum power on each channel is the total transmission power of the home base station and the number of users (that is, the channel number) business.

2.3) After executing step 2.2, check whether the power on each channel of each home base station meets the power of the last time ε is the iteration threshold, if it is not satisfied, set time=time+1, and the algorithm skips to step 2.2 to continue execution; if it is satisfied, the game ends; finally obtain the transmit power set of each channel of the home base station The home base station adjusts the transmit power accordingly.

Based on the two-layer scenario of a home base station network and a cellular network in a layered heterogeneous network, the present invention makes reasonable allocation of power resources and channel resources in the wireless network. To improve the heuristic clustering algorithm based on graph theory, first establish a directed interference graph and calculate the interference weight of each home base station, add the home base station to multiple clusters, and enable the home base station to be able to Using multiple channels improves the spectrum utilization of the system.

The method of the invention uses the principle of game theory to build a non-cooperative game model. The game not only considers the throughput of each home user, but also considers the interference of the home base station transmission power on the macro user and other home users on the channel, which not only improves home The user's throughput ensures the performance of other users.

Beneficial effect

Compared with the prior art, the present invention has the following advantages:

1. Unlike the general heuristic clustering algorithm, the present invention improves the algorithm, so that the home base station can obtain multiple channels by using the algorithm and joint channel allocation.

2. The proposed power game algorithm adjusts the power of each channel individually for each channel of the home base station; the utility function sets a penalty for the interference caused by the transmission power of the home base station to different users, and the penalty for the interference of the macro user Greater than for home users, focus on protecting the performance of macro users.

3. Using graph theory to allocate joint channels between home base stations and macro users, and taking into account cross-layer interference and same-layer interference, the channel division is more reasonable, reducing the interference of home base station clusters and macro users using the same channel, and ensuring the security of macro users It also has a certain improvement effect on the performance of home users.

Description of drawings

Figure 1 is a downlink model diagram of a two-layer network.

Fig. 2 is a schematic flow chart of the method of the present invention.

Fig. 3, Fig. 4, Fig. 5, and Fig. 6 are simulation result diagrams of user outage probability and network throughput in a two-layer network.

detailed description

Below in conjunction with example and description accompanying drawing 1,2 the technical scheme of invention is described in detail:

The present invention utilizes graph theory and game theory to reasonably allocate power resources and channel resources in hierarchical heterogeneous networks: to improve the heuristic clustering algorithm based on graph theory, and first to use graph theory to establish a home base station network Directed interference graph model, calculate the interference weight received by each home base station after joining the cluster, and add the home base station to multiple clusters according to the interference weight, and select the appropriate channel according to the interference received on the cluster after each macro user allocates a channel Cluster pairing, the femtocells in the cluster can use the channels of the paired macro users, so the femtocell can obtain multiple channels; use game theory to establish a non-cooperative game model for the transmit power of each channel of the femtocell, and the game considers the throughput of the home user The amount and the interference caused, after many games, the best set of transmit power is obtained.

The downlink scene of the heterogeneous cellular network studied by the present invention is shown in FIG. 1 , and the overall flow chart of the heterogeneous cellular network resource allocation method based on multi-channel heuristic clustering is shown in FIG. 2 .

A heterogeneous cellular network resource allocation method based on multi-channel heuristic clustering of the present invention comprises the following steps:

1) Joint channel allocation

1.1) Define the home base station set {V}={v ₁ ,v ₂ ,...,v _F }, {v _i } represents home base station i, the number of home base stations is F, the number of macro users and the number of channels equal, both are M, the transmission power of HNB i is P _i , the initial transmission power of all HNBs is P ^f , and the number of channels allocated by HNB i is T _i .

The signal transmitted from the base station to the user's reception will pass through a certain distance and cause fading, and the magnitude of the fading is related to the distance d between the two. The fading formula of the fading model used in this method is as follows:

(1) Fading from macro base station to macro user:

PL(dB)＝30+40*log10(d)+10 (1)

(2) Fading from femtocell to home user, from femtocell to macro user, from macro base station to home user

When d>15,

PL(dB)＝60+25*log10(15)+40*log10(d-15)+10 (2)

When d≤15,

PL(dB)＝40+25*log10(d)+10 (3)

The user's channel gain can be obtained from the channel fading. The channel gain from the macro base station to the macro user m is The channel gain from Femtocell i to macro user m is The channel gain from home base station i to home user j is The range of macro users is m∈{1,2,...,M}, and the range of femtocells is i∈{1,2,...,F}.

Home user i is interfered by home base station j, expressed as follows:

1.2) The home user equipment detects the interference Igi _,j received by itself, and calculates the interference weight from the home base station j. Since it is assumed that the initial transmission power of the home base station is equal to P ^f , the calculation formula of the interference weight of the home base station i by the home base station j is:

Where δ is the threshold of the interference weight, and if it is less than the threshold, it is considered that there is no interference between the home base stations. by formula The sum of the interference weights suffered by the home base station i is calculated to form a set of interference weights {W}={w ₁ , w ₂ , . . . , w _F }.

1.3) Arranging the home base stations in descending order of interference weight to form a new set of home base stations {V'};

1.4) Divide the home base station into clusters {C _n } _n∈M according to the number of channels M.

1.4.1 Take a home base station i from the home base station set {V'} in turn to join cluster C _n , the number of home base stations that cluster C _n will interfere with home base station i is Where {v _i '} is a set of HNBs in the cluster C _n that will interfere with HNB i. W _l ⁿ is the increase of the cluster interference weight after joining the cluster n, and the sum of the added weights after the Femtocell joins the cluster is calculated, and the formula is as follows.

Calculate W _l ⁿ for all clusters to form a weighted set {W _l ⁿ } _n∈M , select the first T _i home base station clusters C _n' with the smallest W _l ⁿ from the set, and add home base station i to it. The HNB is deleted from the base station set {V'}.

1.4.2 Repeat step 1.4.1 until the home base station set {V'} is empty, then the home base station clustering is completed.

1.5) Joint channel allocation between the home base station and the macro user.

1.5.1 The number of macro users is M, and the set of macro users is {Mu}. Arrange the macro users in the order from the closest to the farthest from the macro base station to obtain a new set of macro users {Mu'}, and allocate them to the macro users in sequence a channel;

1.5.2 Take a macro user m from the macro user set {Mu'} in order, and calculate the sum of the interference and noise it receives from the cluster C _n is the channel gain of the interference signal from the home base station i to the macro user m, select a cluster C _n' from the cluster set {C _n } that makes IN _n the smallest, match the macro user m with this cluster, the home base station in the cluster and the macro user m The user uses the same channel, and the home base station cluster and macro user are deleted from the home base station cluster set {C _n } and the macro user set {Mu'};

1.5.3 Repeat step 1.5.2 until both the home base station cluster set {C _n } and the macro user set {Mu'} are empty, and the joint channel allocation is completed.

2) Allocate the transmit power of each channel of the home base station

2.1) According to the channel assignment of the home base station and the macro user obtained in the step 1), the information of each base station is collected at the same time, and the transmission power of each channel of the home base station is played using a non-cooperative model, and a utility function is designed.

The present invention uses a non-cooperative game for the transmission power of the home base station. The number of users in each home base station is random but less than the total number of channels. Each home user uses one channel, and the home base station games the transmission power on each channel. The game is set as follows:

Where I is a set of home base stations, each user uses a channel, is the set of transmit power of each channel of the home base station, T _i is the channel used by the home base station i, is the set of utility functions of the femtocell.

The utility function considers both the throughput of home users and the interference to other users, defined as:

where the first term is the throughput of femtocell i on channel t, is the signal-to-interference-noise ratio of femtocell i on channel t, and the calculation formula is:

C _n is the cluster of femtocells using channel t, is the sum of the interference of other Femtocells in the cluster C _n to the user of Femtocell i on the channel t, P _m is the total transmit power of the macro base station, M is the number of macro users, is the channel gain of the interference signal from the macro base station to the user of the femtocell i, is the interference caused by the macro base station, N0 is the noise power spectral density of the system, B is the channel bandwidth, and N0*B is the noise on the channel. The second term is the penalty for the interference caused by femtocell i on channel t to macro and home users using channel t, and k is the coefficient of the penalty term. The transmission power on each channel is limited, which is: 0<P _i,t ≤ Pmax/T _i , Pmax is the maximum transmission power of the home base station, and the maximum power on each channel is the total transmission power of the home base station and the number of users (ie The quotient of the number of channels).

When the power of the home base station determined by the game does not change, the game reaches the Nash equilibrium. When the Nash equilibrium is reached, no player can rely on changing his strategy alone to obtain better benefits, which can be expressed as:

at this time It is the Nash equilibrium solution of the power game.

2.2) Initialization: set ε=10 ^-5 , time=0, the transmission power of each channel of the home base station is {P _i,t } ^(time) ={Pmax/T _i }, where Pmax is the maximum transmission power of the home base station, T _i is the number of channels used by the home base station i;

2.3) The home base station adjusts the transmit power of each channel according to the formula, the formula is as follows:

where B is the channel bandwidth, k is the coefficient of the penalty term, is the channel gain of the interference signal of the home base station i to the macro user m, is the sum of the interference signal channel gains of home base station i to other home users using channel t, is the SINR of Femtocell i on channel t.

At the same time, the transmit power of each channel is limited, and the formula needs to be included:

Each home base station calculates the transmission power of the home base station according to formulas (11) and (12), and obtains a set of transmission power sets of the home base station.

2.4) After executing step 2.3, check whether the power on each channel of each home base station meets the power of the last time ε is the iteration threshold, if it is not satisfied, set time=time+1, and the algorithm skips to step 2.3 to continue execution; if it is satisfied, the game ends; finally obtain the transmit power set of each channel of the home base station The home base station adjusts the transmit power accordingly.

In summary, the present invention uses graph theory to build a directed interference graph model for the home base station network in the downlink scenario of a hierarchical heterogeneous network, calculates the home base station interference situation according to the interference weight calculation formula, and based on the improved heuristic clustering algorithm Divide the home base stations into multiple clusters, and perform joint channel allocation with macro users. Thereafter, a non-cooperative game model is adopted for the transmission power of the home base station on the occupied channel, and an optimal transmission power set of the home base station is obtained through iterative reaching equilibrium.

As shown in Figure 3 and Figure 4, the outage probability of the heterogeneous cellular network resource allocation method based on multi-channel heuristic clustering is similar to that of the clustering without power control algorithm, and lower than that without clustering and power control , the outage probability of macro users is lower than the other two algorithms; the heterogeneous cellular network resource allocation method based on multi-channel heuristic clustering in Figure 5 and Figure 6 reduces the interference to other users, and the home base station always The throughput is similar to that of the clustering without power control algorithm, but greater than that of the clustering and power control algorithm, and the throughput of the macro base station is better than the other two algorithms.

Claims (1)

1. a kind of isomery cellular network resource distribution method based on the heuristic sub-clustering of multichannel, it is characterised in that including following Step：

1) combined channel distribution

1.1) Home eNodeB set { V }={ v is defined₁,v₂,......,v_F, { v_iRepresent Home eNodeB i, the number of Home eNodeB For F, grand number of users is equal with the number of channel, is M, and the Initial Trans of Home eNodeB are identical, each Home eNodeB transmitting work( Rate is P_i, the channel of Home eNodeB i distribution is T_iIndividual；

1.2) the interference weight that Home eNodeB i is subject to Home eNodeB j is calculatedWhereinIt is Channel gains of the Home eNodeB m to Home eNodeB i user's interference signal,It is Home eNodeB i to this base station user useful signal Channel gain, δ is the thresholding of weight, can be considered as between two Home eNodeB less than this thresholding and not interfere with, by public affairs FormulaThe be disturbed weight sums of Home eNodeB i are obtained, interference weight set { W }={ w is formed₁,w₂,......, w_F}；

1.3) by Home eNodeB by interference weight w_iDescending, forms the new set { V'} after Home eNodeB sequence；

1.4) Home eNodeB is divided by cluster { C according to number of channel M_n}_n∈M：

1.4.1 { take out one family base station i in V'} and add cluster C from Home eNodeB set successively_n, cluster C_nHome eNodeB i can be produced Giving birth to the Home eNodeB number for disturbing is{v_i' it is cluster C_nIn to Home eNodeB i produce interference collection of base stations, cluster C_nIncrease Weight sum beW is chosen from all of cluster_l ⁿMinimum front T_iHome eNodeB is added by individual cluster, { in V'}, leave out the Home eNodeB from Home eNodeB set；

1.4.2 up to Home eNodeB set, { V'} is that sky, then Home eNodeB sub-clustering are completed to repeat step 1.4.1；

1.5) Home eNodeB is distributed with grand user's combined channel：

1.5.1 grand number of users is M, and Hong Yonghuji is combined into { Mu }, by grand user according to apart from macro base station near to remote order Arrangement, obtains newly grand user set { Mu'}, and be followed successively by grand user in order and distribute a channel；

1.5.2 { take out a grand user m in Mu'}, calculate it by cluster C from grand user set in order_nInterference and noise Sum For the channel gain of Home eNodeB i to grand user m interference signals, from gathering Close { C_nIn select one cause IN_nMinimum cluster C_n'Grand user m is matched with the cluster, the Home eNodeB in cluster and grand user Using same channel, the Home eNodeB cluster and grand user are closed { C from Home eNodeB gathering_nAnd grand user set { delete in Mu'} Go；

1.5.3 repeat step 1.5.2 is until Home eNodeB gathering closes { C_nAnd grand user set { Mu'} be empty, and combined channel divides Match somebody with somebody and complete；

2) each channel emission power of distribution Home eNodeB

2.1) initialize：If ε=10^-5, time=0, the transmission power of each channel of Home eNodeB is { P_i,t}^(time)={ Pmax/ T_i, wherein Pmax be Home eNodeB maximum transmission power, T_iFor the number of channel that Home eNodeB i is used；

2.2) each channel emission power of Home eNodeB is calculated Wherein B is channel width, and k is the coefficient of penalty term,For channel gains of the Home eNodeB i to grand user m interference signals,For Home eNodeB i to other using channel t domestic consumer interference signal channel gain sum,It is Home eNodeB Signal to Interference plus Noise Ratio of the i on channel t, computing formula isC_nIt is using channel t Home eNodeB cluster,For cluster C_nIn other Home eNodeB on channel t to the interference of Home eNodeB i user it With P_mIt is macro base station total transmission power, M is grand number of users,Interference signal for macro base station to the user of Home eNodeB i is believed Road gain,It is interference that macro base station is caused, N0 is the noise power spectral density of system, B is channel width, N0*B is Noise on channel.On each channel, transmission power is restricted, is 0 ＜ P_i,t≤Pmax/T_i, Pmax is that Home eNodeB is maximum Transmission power, the peak power on each channel are the business of Home eNodeB total transmission power and number of users (i.e. the number of channel)；

2.3), after execution step 2.2, check whether the power on each each channel of Home eNodeB is met with the power of last timeε is iteration threshold；If be unsatisfactory for, time=time+1, algorithm is made to jump to step 2.2 continue executing with；If it is satisfied, then game terminates；Home eNodeB each channel emission power set is finally obtained Home eNodeB adjusts transmission power accordingly.