Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Low complexity hard output maximum likelihood detection algorithm for LTE and WiFi MIMO

IP.com Disclosure Number: IPCOM000241179D
Publication Date: 2015-Apr-01
Document File: 2 page(s) / 14K

Publishing Venue

The IP.com Prior Art Database

Related People

Raul Casas: INVENTOR [+4]

Abstract

The proposed algorithm only has a 0.5 dB performance loss compared to the full search ML detection at 12dB SNR. And the complexity counted for the expansion is just about N+K*B, where N is the QAM size, K is the K best value described in step 5 of the previous section. B is the number of first neighbors of a constellation node(range from 4 to 9).

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 58% of the total text.

Page 01 of 2

     Cadence Design Systems, Inc. Inventors: Lian Huai, Samer Hijazi, Raul Casas

TITLE

Low complexity hard output maximum likelihood detection algorithm for LTE and WiFi MIMO

PROBLEM AND SOLUTION

MIMO has been incorporated in modern wireless communication standards such as IEEE 802.11n (Wi-Fi), 4G, 3GPP LTE, LTE-Advanced, WiMAX and HSPA+ due to its excellence in spectral efficiency. There are lots of detection algorithms for MIMO at the receiver side. The full search maximum likelihood detection has the optimal bit error rate. But the computational complexity makes it unfeasible for hardware implementation. Some other algorithms like ZF or K-best detection, are much lower in complexity but at the cost of performance. Therefore, it is still a challenge to realize efficient MIMO communication systems due to the resulted area and power complexity. It is important to design a low power reduced complexity MIMO detection algorithm at the receiver end which has a near optimal performance.

In order to obtain near optimal performance with the least computation complexity in MIMO detection, we proposed a novel way which adopted some new ideals along with the K best ideal.

For example in our 4x4 channel model, we have four symbols s1, s2, s3 and s4 to be detected for each sub-carrier.

1. We first reorder the columns of the channel matrix to make the channel column powers in a decreasing order.

2. Then we apply QR decomposition to make it possible to search from last layer (weak...