MIMO信号检测系统量子算法的优化及运用英文小论文
Abstract
The optimal solution of signal detection is a NP (Nondeterministic Polynomial) problem. Aimed at the problems that neural network is prone to the local optimum and simple genetic algorithm has the shortcoming of slow convergence, a new type of algorithm optimized by quantum is proposed and applied into the MIMO/MIMO-OFDM detection systems : It makes use of Quantum Genetic Algorithm（QGA）to optimize the initial data of neural network. In this scheme, the output of detector by the QGA as the input of detector by neural network to avoid the bit -error rate for selecting initial data randomly and improve further the detection property. Simulation results show the proposed method is good for the improvement of the detection rate and reduction of bit-error rate.
1. Introduction
Quantum computing is an emerging computing model, which combines with quantum theory、information theory and computer science. it makes use of the superposition、quantum parallel、quantum entanglement and other properties of quantum systems to achieve more efficient than the classic computing model calculation[1]. Because quantum properties have a unique function in information field, it may exceed the existing limit of classical information systems in case of improving processing speed、ensuring information security、increasing capacity and improving accuracy of information and so on. So it is very important significant to apply quantum computation to information processing research. Current research topics include: quantum computers、quantum communication and quantum cryptography which have made a major breakthrough in theory and experiment. A new type of quantum optimization algorithm combines with quantum computation、genetic algorithms and neural networks is proposed in this paper and applied into signal detection of MIMO (Multiple-Input Multiple-Output) and MIMO-OFDM (Multiple-Input Multiple -Output Orthogonal Frequency Division Multiplexing) systems.
Genetic algorithm is a mechanism algorithm which simulates the natural evolution of species. However, the optimal solution for some problems is difficult to find by classic GA, it makes people try to research more efficient and fast quantum genetic algorithm which combines with quantum theory and genetic algorithm [2]. The proposed QGA in this paper has more parallel processing power and faster convergence speed than the classical genetic algorithm, for
it uses quantum parallel、quantum entanglement、multi-state gene encoding qubits、quantum revolving door update and quantum crossover operation. Literature [3] shows that the CDMA multi-user detection which based on QGA has a higher detection efficiency than the GA and other traditional signal detection algorithms.
Neural network can be used in the field of signal detection as it has advantages of information distributed storage、large-scale parallel processing and highly adaptive fault tolerance, etc. RBF(Complex-valued Radial Basis Function) Neural Network is a kind of nonlinear signal processing technique, it has excellent features of learning speed、network structure determined adaptively and output has nothing to do with the initial weights, etc. Literatures [4,5] has been under close to optimal Bayesian detection performance by neural network in CDMA system. This paper attempts to study neural networks combined with QGA, and obtain better detection performance in MIMO and MIMO-OFDM systems based on neural network which optimized by QGA.
2. Quantum Genetic Algorithm (QGA)
2.1 Quantum bit (qubit)
The smallest unit of information stored in a two-state quantum computer is called a quantum bit or qubit. A qubit may be in the“1”state, in the“0”state, or in any linear superposition of the two. The state of a qubit can be represented as
= （1）Where, and are complex numbers, .
A Q-bit individual as a string of m Q-bits is defined as （2）
where ， .
2.2 Mechanism of the QGA
QGA is a probabilistic algorithm which similar to other evolutionary algorithms.
However, QGA maintains a population of Q-bit individuals, , at generation t, where n is the size of population, and is a Q-bit individual defined as （3）
Where, is the number of Q-bits, i.e., the string length of the Q-bit individual, and . The procedure of QGA is described as follows:
Step1：initialization ；
Step2：make by observing the states of ；
Step3：evaluate by observing the states of and obtained the best fitness of the individual as the target of next evolution values；
Step4：while (not termination-condition) do
begin ；
Step5：update using Q-cross；
Step6：update using Q-gates, return Step2。
2.3 Algorithm performance testing
The proposed quantum genetic algorithm represents chromosomes with quantum bit code, completes the evolutionary search with quantum crossover and quantum revolving gates. In order to verify the feasibility and effectiveness of the proposed quantum genetic algorithm and compare with the classical genetic algorithm by following two typical complex function.
(1) De Jong function1124