CN111082811A - Low-power-consumption Bluetooth Viterbi joint demodulation decoding algorithm with S =8 coding mode - Google Patents

Abstract

The invention belongs to the technical field of demodulation and decoding of wireless communication, and particularly relates to a low-power-consumption Bluetooth Viterbi joint demodulation and decoding algorithm in an S-8 coding mode. Aiming at an S-8 coding mode, a Gaussian Frequency Shift Keying (GFSK) process and a low-power consumption Bluetooth coding process are modeled into a finite state machine, and the model is subjected to joint demodulation decoding by utilizing a Viterbi decoding algorithm. The invention considers the practical application scene of the low-power-consumption Bluetooth, has low algorithm complexity, can obviously reduce the error rate and improve the sensitivity of the receiver.

Description

A Low Energy Bluetooth Viterbi Joint Demodulation and Decoding Algorithm Based on S=8 Coding

technical field

The invention belongs to the technical field of demodulation and decoding of wireless communication, and in particular relates to a low-power consumption bluetooth Viterbi joint demodulation and decoding algorithm with S=8 coding mode.

Background technique

Bluetooth low energy technology is one of the representative technologies of the Internet of Things, which is widely used in computing devices with low cost and low power consumption, as well as in short-range wireless communication scenarios with low data rates and low duty cycles. With the development of the Internet of Things, in the Bluetooth protocol version 5.0 in 2016, Bluetooth Low Energy added a new coding physical layer and two coding schemes (S=2, S=8), corresponding to the information transmission rate of 500kb/s and 125kb/s. The coding physical layer enhances the stability of Bluetooth signal transmission. The low-power Bluetooth signal transmission distance can be increased by up to 4 times without increasing the transmission power, which greatly expands the application field and development prospects of low-power Bluetooth in the Internet of Things.

For the S=8 coding physical layer of low-power bluetooth, the receiver uses Viterbi joint demodulation and decoding, which will greatly improve the sensitivity of the receiver, which is a very practical problem.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a low-power consumption bluetooth Viterbi joint demodulation and decoding algorithm with low computational complexity and low error rate.

The core of the present invention is to model the Gaussian frequency shift keying modulation (GFSK) process and the low-power bluetooth coding process as a finite state machine, and jointly perform demodulation and decoding according to the Viterbi decoding algorithm. According to the S=8 coding mode, the present invention uses the Viterbi decoding algorithm to jointly demodulate and decode this model. The present invention takes into account the actual application scenarios of low-power bluetooth, has low algorithm complexity, and can significantly reduce the bit error rate and improve receiver sensitivity.

The low power consumption Bluetooth Viterbi joint demodulation and decoding algorithm of the S=8 coding mode provided by the present invention has the following specific steps.

The first step, for a bluetooth signal r(t) sent through S=8 encoder and modulator, its phase is:

From the GFSK signal model, there are:

Q()为高斯Q函数，I_k∈{±1}是输入的随机比特流，h是GFSK调制指数，通常取为0.5；BT为时间-带宽乘积，一般取为0.5，脉冲成型函数q(t)的有效持续时间为2T。in,

Q() is the Gaussian Q function, I _k ∈{±1} is the input random bit stream, h is the GFSK modulation index, usually taken as 0.5; BT is the time-bandwidth product, generally taken as 0.5, the pulse shaping function q( The effective duration of t) is 2T.

当t<0和t≥2T时，g(t)＝0；Therefore, there are: when t<0, q(t)=0, when t≥2T,

When t<0 and t≥2T, g(t)=0;

According to the Bluetooth low energy 5.0 protocol, the bit stream is mapped to '1100' and '0' to '0011' after encoding and mapping extension, so the '+1' and '-1' in the I _k sequence ' appear in pairs, so the above formula can be rewritten as:

因此The first item in the above formula,

therefore

In the above formula, I _k (k=n...n+7) is only determined by the output of the convolutional code encoder, so the finite state machine that can combine the coding and modulation in the S=8 mode can be equivalent to bit stream coding (convolutional code). code) finite state machine.

In the second step, according to the finite state machine modeled by the received signal in the first step, it is decoded according to the Viterbi decoding method to achieve joint demodulation and decoding.

The Viterbi joint demodulation and decoding algorithm used in the second step is as follows:

(1) Calculate the branch metric according to the output of each state of the finite state machine that the received signal is modeled into in the first step;

(2) Calculate the path metric value according to the transition relationship of each state of the finite state machine;

(3) For each state, retain the arrival path (survivor path) with the smallest path metric value calculated in the previous step, and delete the rest of the arrival paths;

(4) Find the surviving path corresponding to the minimum path metric value, and output the decoded bits in turn according to the state transition relationship of the finite state machine.

状态由I_k(k＝n,n+1…n+7)的取值决定，由于I_k是由卷积码编码器的输出的两个比特经过映射延展得到的，I_k的8个比特的序列仅由序列“0011”和“1100”组合而成，共4种组合情况；因此维特比译码算法在计算分支度量时，计算在nT<t<(n+8)T时刻内，接收信号r(t)和包含所有

状态的

(共4种状态组合，i＝0…3)的相关度量，并保留具有最大相关度量值的路径，即：Due to the received Bluetooth signal, its phase is shown in equation (6);

The state is determined by the value of I _k (k=n, n+1...n+7). Since I _k is obtained by mapping and extending the two bits of the output of the convolutional code encoder, the 8 bits of I _k The sequence is only composed of the sequences "0011" and "1100", and there are 4 combinations in total; therefore, when the Viterbi decoding algorithm calculates the branch metric, it is calculated at the moment of nT<t<(n+8)T, and the receiving The signal r(t) and contains all

state

(a total of 4 state combinations, i=0...3), and keep the path with the largest correlation metric value, namely:

In the present invention, the relationship between the input, output and state transition of the coded modulation joint finite state machine in the S=8 coding mode is shown in Table 1.

Advantages of the method of the invention

(1) Compared with the traditional decoding method of decoding after demodulation, the joint Viterbi demodulation and decoding algorithm significantly reduces the bit error rate and improves the receiver sensitivity;

(2) Considering the limited computing power and low power consumption of the low-power Bluetooth receiver, the algorithm has low complexity and is easy to implement.

Description of drawings

FIG. 1 is a state transition diagram of the code-modulation combination in the S=8 coding mode.

FIG. 2 is a comparison of the bit error rate performance of the joint demodulation and decoding algorithm under the S=8 coding scheme.

Detailed ways

The present invention will be further described below through specific embodiments.

As an embodiment, the present invention uses a computer to simulate the complete flow of Bluetooth GFSK signal encoding-modulation-demodulation-decoding. During the simulation process, a 1000-bit data packet is randomly generated, and the GFSK modulation index is 0.5. The coding method specified by the Bluetooth low energy protocol is adopted, and 1000 Monte Carlo experiments are carried out under the Bluetooth low energy S=8 coding scheme. The final bit error rate performance and comparison are shown in Figure 2, where the x-axis is the simulated signal-to-noise ratio, and the y-axis is the bit error rate after decoding at the receiving end.

The differential hard demodulation decoding algorithm in the figure (the line with a diamond on the upper side) corresponds to directly taking the optimal phase difference of the received signal and making a hard decision as 0/1 bits, and then demapping and inputting it to the Viterbi decoder for hard decoding; differential The soft demodulation and soft decoding algorithm (the line with the triangle on the upper side) corresponds to the input of the Viterbi soft decoding after the demodulation phase is demapped; the Viterbi hard demodulation and hard decoding algorithm (the line with the cross on the lower side) corresponds to the The received signal is first demodulated to 0/1 bits by the Viterbi algorithm, and then de-mapped and input to the Viterbi decoder for hard decoding; the Viterbi joint demodulation and decoding algorithm (the line with dots on the lower side) corresponds to the algorithm of the present invention. It can be seen that the bit error rate performance of the joint demodulation and decoding algorithm proposed by the present invention is improved by 4dB to 6dB compared with the traditional soft decoding, and is improved by 2dB to 3dB compared with the Viterbi hard demodulation and hard decoding.

Table 1, the relationship between the input, output and state transition of the coded modulation joint finite state machine in S=8 coding mode

references

[1]Bo Y, Yang L, Chong C C.Optimized Differential GFSK Demodulator[J].IEEE Transactions on Communications,2011,59(6):1497-1501.

[2] Anderson J B, Aulin T, Sundberg C E. Digital Phase Modulation [J]. Applications of Communications Theory, 1986: 412-412.

Claims (3)

1. a low power consumption bluetooth Viterbi joint demodulation and decoding algorithm of S=8 coding mode, is characterized in that, concrete steps are as follows: The first step, for a bluetooth signal r(t) sent through S=8 encoder and modulator, its phase is:

From the GFSK signal model, there are:

q(t)=∫ ₀ ^t g(τ)dτ (3) in,

Q() is the Gaussian Q function, I _k ∈{±1} is the input random bit stream, h is the GFSK modulation index; BT is the time-bandwidth product, and the effective duration of the pulse shaping function q(t) is 2T; Therefore, there are: when t<0, q(t)=0, when t≥2T,

When t<0 and t≥2T, g(t)=0;

According to the Bluetooth low energy 5.0 protocol, after the bit stream is encoded in S=8 mode and is mapped and extended, the bit '1' is mapped to '1100', the bit '0' is mapped to '0011', and the '+1' in the I _k sequence and '-1' appear in pairs, so the above formula is rewritten as:

Since the first term in the above formula,

Therefore:

In the above formula (6), I _k (k=n...n+7) is only determined by the output of the convolutional code encoder, so the finite state machine of the coding and modulation joint in the S=8 mode is equivalent to bit stream coding That is, the finite state machine of the convolutional code; In the second step, the received signal is jointly demodulated and decoded according to the Viterbi decoding algorithm according to the finite state machine modeled by the received signal in the first step. 2. joint demodulation and decoding algorithm according to claim 1, is characterized in that, in the second step, described utilizes Viterbi decoding algorithm to carry out joint demodulation and decoding to received signal, and concrete process flow is as follows: (1) Calculate the branch metric according to the output of each state of the finite state machine that the received signal is modeled into in the first step; (2) Calculate the path metric value according to the transition relationship of each state of the finite state machine; (3) For each state, keep the arrival path with the smallest path metric value calculated in the previous step, survive the path, and delete the rest of the arrival paths; (4) Find the surviving path corresponding to the minimum path metric value, and output the decoded bits in turn according to the state transition relationship of the finite state machine. 3. joint demodulation decoding algorithm according to claim 2, is characterized in that, due to the received bluetooth signal, its phase is:

The state is determined by the value of I _k (k=n, n+1...n+7). Since I _k is obtained by mapping and extending the two bits of the output of the convolutional code encoder, the 8 bits of I _k The sequence is only composed of the sequences "0011" and "1100", and there are 4 combinations in total; therefore, when the Viterbi decoding algorithm calculates the branch metric, it is calculated at the moment of nT<t<(n+8)T, and the receiving The signal r(t) and contains all

state

Relevance metrics for a total of 4 state combinations, i=0...3; and keep the path with the largest relevant metric value, namely:

Images