Title :
On Integer-Valued Almost-Perfect Sequences
Author_Institution :
Dept. of Electr. Eng., Nat. Sun Yat-sen Univ., Kaohsiung, Taiwan
Abstract :
In this letter, a new class of integer-valued sequences with almost-perfect periodic autocorrelation function is proposed. To construct these sequences, two groups of base sequences with two and four base sequences in each group are developed, and each base sequence is generated from a generalized Hadamard matrix. The new constructed sequences are the sum of the linear combinations of base sequences. Compared with existing almost-perfect sequences whose elements are complex floating-point numbers, integer-valued almost-perfect sequences have the advantages of error-free quantization and less memory space for implementation. In addition, the energy efficiency of a sequence is derived from theoretical expressions. The optimality criteria for coefficients are also obtained to maximize energy efficiency.
Keywords :
Hadamard matrices; binary sequences; correlation methods; energy conservation; floating point arithmetic; quantisation (signal); telecommunication power management; Hadamard matrix; base sequence linear combination; energy efficiency maximization; error-free quantization; floating-point number; integer-valued almost-perfect sequence; periodic autocorrelation function; Computers; Correlation; Memory management; Peak to average power ratio; Quantization (signal); Systematics; Vectors; Hadamard matrix; Integer-valued; base sequences; integer-valued; periodic autocorrelation function; periodic autocorrelation function (PACFs);
Journal_Title :
Communications Letters, IEEE
DOI :
10.1109/LCOMM.2014.2381228
