CN111601360A - A relay selection method in a millimeter wave communication system - Google Patents

Abstract

The invention belongs to the technical field of millimeter wave communication, and in particular relates to a relay selection method in a millimeter wave communication system. The technical scheme of the present invention is to add a new load factor broadcast as a part of the system information on the basis of the existing relay search and selection scheme based on received power, and equip the network with a powerful access control method. Helps load balancing and effectively utilizes millimeter-wave spectrum resources. Secondly, the present invention obtains the relay search and selection strategy based on the optimal stopping theory, and optimizes the connection threshold, so that the relay search and selection of the millimeter wave communication using this strategy have the optimal throughput.

Description

A relay selection method in a millimeter wave communication system

technical field

The invention belongs to the technical field of millimeter wave communication, and in particular relates to a relay selection method in a millimeter wave communication system.

Background technique

The multi-relay mmWave communication network is shown in Figure 1. As a prominent feature in 5G networks, MmWave makes direct communication between UEs easy to be blocked. When the direct communication between UEs cannot be carried out, the communication can be carried out by means of a relay. The high-density relay deployment scheme in the 5G network makes a large number of available relays around the UE. The UE needs to perform basic relay search and selection before communicating via the relay. Relay search generally consists of two steps: (1) searching and acquiring synchronization with the relay, (2) decoding system information containing basic information for accessing the relay. In relay search, the UE shall search for the relay with the strongest received power. Once the UE finds a suitable relay, it can select that relay to continue communication. Relay selection based on received power In multi-layer heterogeneous networks, it may cause load imbalance problems in the network. Furthermore, MmWave communication utilizes a large amount of spectrum resources in the MmWave frequency band (30-300GHz) to achieve multi-Gbps data rates. In 5G networks, the problem of load imbalance may become more serious, which will make the rich wireless resources brought by mmWave spectrum underutilized.

For the traditional relay search and selection technology, the scheme of selecting the maximum received power as the communication relay is usually adopted. The relay broadcasts synchronization signals and system information, and the synchronization signal period is T _syn seconds. There may be L transmit-receive spatial signature pairs during synchronization. It is assumed that the UE learns the beamforming direction according to the detection of the relay synchronization signal. Suppose there are 30 relays around the UE, and these relays may have established connections with a large number of active UEs.

As shown in Figure 2, the specific steps for relay search and selection based on the maximum received power are as follows:

First, the UE checks the first relay, obtains synchronization and determines the beamforming direction after the beamforming scan period LT _syn time. The UE then measures the corresponding SNR received from the relay.

If the received SNR measured from the current relay is below a certain SNR threshold Γ, the UE stops checking the current relay and starts checking another relay. Otherwise, the UE records the received power of the current relay.

Then, the UE checks each relay, and records the received power for the relay that satisfies the minimum communication signal-to-noise ratio until the 30th relay is checked.

Finally, the UE selects the relay with the largest recorded received power for random access access and starts data communication.

The problems existing in this traditional method include: First, when the UE searches and selects the relay, it does not consider the number of active UEs connected to the relay, which will cause a load imbalance problem in the communication network, making the power low but not enough. The number of micro-relay connections that also work in the millimeter-wave spectrum is extremely small, resulting in an unbalanced load on the millimeter-wave network and wasting millimeter-wave spectrum resources. Secondly, in each communication cycle of the UE, all relays around the UE will be checked for relay search and selection, which will cause the UE to waste a lot of time in checking the relays, reduce the effective communication time in the communication cycle, and reduce the throughput. amount decreased. Finally, due to factors such as time-varying radio environment, UE mobility and clock drift, UE synchronization with previously checked relays may become outdated, resulting in failure of communication access.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a relay selection method based on load sensing and optimal stopping theory in view of the above problems, which can overcome the above problems.

The technical scheme of the present invention is: a method for selecting a relay in a millimeter wave communication system, the millimeter wave communication system includes a source user, a plurality of relays and a sink user, wherein the relay is used for signal transmission and reception, and the source user It is used for signal transmission, and the sink user is used for signal reception; it specifically includes the following steps:

S1, the source user detects the synchronization signal broadcast by the i-th relay, and the synchronization signal includes relay load information, and measures whether the SNR _i received from the relay i is greater than the set threshold Γ, if so, then enter step S2, otherwise, Select the jth relay, i≠j, update i=j, repeat step S1;

S2. The source user extracts the load information of the relay i from the synchronization signal, and obtains the expected scheduling probability β _i :

β _i =1/(M _i +1)

Wherein, M _i is the number of active users served by relay i;

S3. The source user calculates the selection metric R _{i of the relay i} :

R _i =β _i log(1+SNR _i )

S4, determine whether R _i is greater than or equal to the connection threshold μ, if so, go to step S5, otherwise, the source user selects the jth relay, i≠j, update i=j, go back to step S1;

The calculation formula of the connection threshold μ is:

Among them, W is the channel bandwidth, λ is the optimal throughput, the calculation method is: set the initial value of the optimal throughput λ ^* (0) = k, k is any value greater than 0, iterative calculation:

为独立同分布随机变量

其累积分布函数为

累积分布函数

中的x即为

的可能取值。每个中继SNR不低于阈值Γ的概率都相同，且为Ρ，

T_syn为中继广播同步信号的周期，T_sib为同步信号中包含中继负荷信息的系统信息块的周期；迭代至收敛后，取收敛值为最优吞吐量λ^*；t is the number of iterative calculation rounds, and the initial value is 0. T _ra is the fixed time required for the random access process, T _data is the data transmission duration of the communication between the source user and the relay, assuming the expected selection metric of the relay

is an independent and identically distributed random variable

Its cumulative distribution function is

cumulative distribution function

The x in is

possible values. The probability that the SNR of each relay is not lower than the threshold Γ is the same, and is P,

T _syn is the period of the relay broadcast synchronization signal, and T _sib is the period of the system information block containing the relay load information in the synchronization signal; after iterating to convergence, take the convergence value as the optimal throughput λ ^* ;

S5. The source user selects the relay i, and the source user and the sink user establish a connection with the selected relay i to perform data communication.

The technical scheme of the present invention is to add a new load factor broadcast as a part of the system information on the basis of the existing relay search and selection scheme based on received power, and equip the network with a powerful access control method. Helps load balancing and effectively utilizes millimeter-wave spectrum resources. Secondly, the present invention obtains the relay search and selection strategy based on the optimal stopping theory, and optimizes the connection threshold, so that the relay search and selection of the millimeter wave communication using this strategy have the optimal throughput.

The beneficial effects of the present invention are:

(1) The present invention adds a new load factor to the system information broadcast by the candidate relay: expected scheduling probability. By incorporating the expected scheduling probability into the selection metric, the present invention provides a powerful access control method for the network, and obtains better network load balancing characteristics.

(2) The present invention converts the relay search and selection scheme performance into a throughput optimization problem, and solves the problem by using the optimal stopping theory, and obtains the connection threshold with the optimal throughput. Combined with the connection threshold based on the optimal stopping theory, the relay search and selection scheme proposed by the present invention has the optimal throughput in the millimeter wave relay selection.

(3) Since the threshold value and connection strategy calculated by adopting the optimal stopping strategy are only related to the currently checked relay, the previously checked relay no longer affects the final communication connection, which makes the present invention avoid time-dependent Due to factors such as changing radio environment, UE mobility, and clock drift, the synchronization of the UE with the previous relay or the extracted relay load information may become outdated, resulting in a situation where the communication fails.

Description of drawings

Figure 1 is a schematic diagram of a multi-relay millimeter-wave communication system;

Fig. 2 is the relay search and selection flow chart based on maximum received power;

Figure 3 is a schematic diagram of two-hop millimeter wave relay communication;

Fig. 4 is the relay search and selection flow chart of the present invention;

Figure 5 is a schematic diagram of load distribution under different associations in a two-layer network, wherein (a) is the maximum received power association; (b) is the maximum signal-to-noise ratio association; (c) is the maximum selection metric association;

FIG. 6 is a schematic diagram of throughput performance comparison under different relay search and selection strategies.

Detailed ways

The present invention is mainly applied to a 5G millimeter wave communication system. The application scenario is two-hop millimeter-wave communication between two UEs by means of a relay. As shown in Figure 3, there are a large number of relays around the UE where the two-hop mmWave relay communication is performed. The first hop between the source UE and the relay is millimeter wave communication, and the source UE needs to perform relay search and selection before data communication. The second hop between the relay and the sink UE is reliable transmission. The present invention focuses on proper relay search and selection in the first hop.

The network elements involved in the present invention include: a relay and a user (UE). In the 5G network two-hop millimeter wave relay communication scheme, the relay is used for signal transmission and reception, the source user is used for signal transmission, and the sink user is used for signal reception.

The present invention designs a relay search and selection technology based on load sensing and optimal stopping theory. In this technology, the expected scheduling probability β _i of the relay is used as the load factor in the relay system information, and the relay broadcasts it to the UE. The UE uses the optimal stopping theory to calculate the connection threshold μ, and uses the load factor and the relay SNR to calculate the selection metric R _i .

As shown in FIG. 4 , it is a schematic diagram of the relay search and selection process of the present invention. The method scene parameters of the present invention are as follows:

1. The candidate relay periodically broadcasts a synchronization signal with a period of T _syn seconds. It is assumed that the UE learns the beamforming direction by detecting the synchronization signal broadcast by the relay. There may be L transmit-receive spatial signature pairs during synchronization.

2. The candidate relay periodically broadcasts a system information block containing relay load information with a period of T _sib seconds. The relay load information is located in the first system information block. It is assumed that each UE can decode the system information block and extract the relay load information in one go.

3. Assume that the number of active UEs served by relay i is M _i .

4. Assume that the random access process takes a fixed time T _ra seconds. It is assumed that the data transmission duration between the source UE and the data communication relay is T _data seconds.

Relay search and selection steps:

Step 1. The UE checks relay i, obtains synchronization after the beamforming scan period LT _syn time, and determines the beamforming direction. The UE then measures the corresponding SNR received from the relay.

• If the received SNR (denoted as SNR _i ) measured from relay i is below a certain threshold Γ, the UE stops checking the current relay i and starts checking another relay j.

· If SNR _i ≥ Γ, go to step 2.

Step 2. The UE decodes the first system information block to extract the load information of relay i, that is, the expected scheduling probability β _i , β _i =1/(M _i +1).

Step 3. The UE calculates the selection metric Ri for relay _i , where Ri = β _i log(1+SNR _{i )} .

Step 4. The UE compares the selection metric R _i corresponding to the current search relay i with the pre-calculated connection threshold μ.

· If R _i ≥ μ, go to step 5.

· If R _i < μ, the UE selects another relay j, proceeds to step 1, and repeats the above steps.

Step 5. The UE completes the relay search and selection, and sends a random access preamble to the currently searched relay i to start the random access procedure. Thus, a connection is established between the UE and the selected relay, after which data communication can be scheduled.

The specific steps for the UE to pre-calculate the connection threshold μ are as follows:

为独立同分布随机变量

其累积分布函数为

假设每个中继SNR不低于阈值Γ的概率都相同，且为Ρ(SNR≥Γ)。Hypothetical relay's expected choice metric

is an independent and identically distributed random variable

Its cumulative distribution function is

It is assumed that the probability that the SNR of each relay is not lower than the threshold Γ is the same, and is P(SNR≥Γ).

Step 1. Set the optimal throughput λ ^* initial value λ ^* (0)=k, where k is any value greater than 0.

Step 2. Iterative calculation

迭代至收敛，取其收敛值为最优吞吐量λ^*。in

Iterate until convergence, and take its convergence value as the optimal throughput λ ^* .

Step 3. When the channel bandwidth is W, the connection threshold

By adding a load factor to the system information broadcast by the relay, the load balancing effect obtained is shown in Figure 5. Figure 5 shows the implementation of load distribution under different association schemes in a two-layer network. In the network, there is one large relay, four micro relays, and a large number of UEs are evenly distributed. Figure 5(a), (b) are the maximum received power correlation and the maximum signal-to-noise ratio correlation, respectively. Compared with the distribution of serving UEs of each relay in the maximum selection metric correlation in Figure 5(c), it is obvious that the maximum selection metric The network under the association scheme has better load balancing. This is because the maximum metric association scheme takes the relay load factor into consideration in the relay search and selection process, so the scheme of the present invention is obviously effective.

Second, as shown in Fig. 6, the relay search and selection strategy of the maximum selection metric considers both received power and load, and the throughput performance is better than other relays when the proposed selection metric is combined with the derived optimal stopping rule Search and select strategies.

Claims (1)

1. A relay selection method in a millimeter wave communication system, the millimeter wave communication system comprising a source user, a plurality of relays and a sink user, wherein the relay is used for signal transmission and reception, and the source user is used for signal transmission, The sink user is used for signal reception; it is characterized in that, comprises the following steps: S1, the source user detects the synchronization signal broadcast by the i-th relay, and the synchronization signal includes relay load information, and measures whether the SNR _i received from the relay i is greater than the set threshold Γ, if so, then enter step S2, otherwise, Select the jth relay, i≠j, update i=j, repeat step S1; S2. The source user extracts the load information of the relay i from the synchronization signal, and obtains the expected scheduling probability β _i : β _i =1/(M _i +1) Wherein, M _i is the number of active users served by relay i; S3. The source user calculates the selection metric R _{i of the relay i} : R _i =β _i log(1+SNR _i ) S4, determine whether R _i is greater than or equal to the connection threshold μ, if so, go to step S5, otherwise, the source user selects the jth relay, i≠j, update i=j, go back to step S1; The calculation formula of the connection threshold μ is:

Among them, W is the channel bandwidth, λ is the optimal throughput, the calculation method is: set the initial value of the optimal throughput λ ^* (0) = k, k is any value greater than 0, iterative calculation:

t is the number of iterative calculation rounds, the initial value is 0, T _ra is the fixed time required for the random access process, T _data is the data transmission duration of the communication between the source user and the relay, assuming the relay's desired selection metric

is an independent and identically distributed random variable

Its cumulative distribution function is

The probability that the SNR of each relay is not lower than the threshold Γ is the same, and is P,

T _syn is the period of the relay broadcast synchronization signal, and T _sib is the period of the system information block containing the relay load information in the synchronization signal; after iterating to convergence, take the convergence value as the optimal throughput λ ^* ; S5. The source user selects the relay i, and the source user and the sink user establish a connection with the selected relay i to perform data communication.

Images