Search

In this article, a capacity-oriented envelope correlation coefficient (ECC) that incorporates a preferred channel model for assessing a multiple antenna system of a multiple-input-multiple-output (MIMO) mobile device is proposed. Originated from the rudimentary law of ergodic channel capacity, the ECC for multiple antennas (ECC-M) is proven to holistically reflect the impact of radiation patterns and total efficiencies of multiple antennas on channel capacity. The closed-form ECC-M expressions for Rayleigh fading channel model and ray-based spatial channel model are studied, leading to an easy-to-use ECC-M expression that approximately incorporates a realistic channel model. The correlation of the proposed ECC-M to the channel capacity is demonstrated through the Monte Carlo (MC) simulation of three four-element MIMO antenna testbeds in two channel models, showing that the antenna total efficiency and a realistic channel model involved in the ECC-M play an important role in assessing the performance of multiple antennas. It is justified that the difference in channel capacities of two multiple antenna systems is highly relevant to their ECC-Ms at high signal-to-noise ratios (SNRs). The proposed ECC-M is also experimentally validated through over-the-air (OTA) throughput measurement, revealing that the ECC-M can effectively assess the OTA performance of multiple antenna systems for MIMO terminal devices. [ABSTRACT FROM AUTHOR]

A cost-effective and accurate method for predicting over-the-air (OTA) data throughput of multiple-input and multiple-output (MIMO) wireless devices in the antenna design stage is proposed. The method incorporates passive characteristics of MIMO antennas, realistic channel model, and conductive measurement results in a computer simulation framework. With extracted relation between received power and signal-to-noise ratio (SNR) of a wireless device at antenna ports by a conductive test and measured passive characteristics of MIMO antennas, OTA throughput versus received power in a realistic channel model is obtained by the Monte Carlo simulation of Shannon’s law. The unique features of this method include: no channel emulator equipment is required; a wide variety of precoding schemes can be applied; passive characteristics of antennas are fully incorporated; and frequency-selective fading channel can be adopted. The method has been intensively validated by comparing the calculated OTA throughputs with those measured by a multiprobe anechoic chamber method under spatial channel model extension (SCME) in three aspects: spatial correlation, radiation efficiency, and mutual coupling, showing as low as 0.2–0.3 dB discrepancy in power level in most cases. The method can be used to evaluate a MIMO array antenna design of a wireless device with respect to OTA performance. [ABSTRACT FROM AUTHOR]

The self-curing decoupling technique is substantially extended to apply to multi-input multi-output (MIMO) antenna arrays in mobile terminals. Using this proposed technique, the mutual coupling between antennas can be reduced by inlaying lumped element capacitors on the ground plane. The decoupling method does not require any physical connection or obstruction between coupled antennas. The significance of the proposed decoupling technique is fourfold: 1) it works for inverted-F antennas (IFAs), monopole antennas, and loop antennas, the most three commonly used antenna forms for mobile terminals; 2) the inherent improvement on matching conditions after decoupling; 3) the flexibility in implementation owing to its detachment from the antenna body; and 4) the applicability to an array with more than two elements. Four demonstrating examples, including two IFAs, two monopole antennas, two loop antennas, and a MIMO array with four elements of different antenna types are studied by EM modeling, passive tests, and MIMO over-the-air (OTA) measurements. The experimental results show that significant improvements in port isolation (from 9 to 30 dB), total efficiency (from 53% to 69%), and above all, the system data throughput can be achieved. The working mechanism of the method is revealed and justified through the partial element equivalent circuit (PEEC) model in the fifth example. [ABSTRACT FROM AUTHOR]