Marilynn P. Wylie-Green

Throughput, Capacity, Handover and Latency Performance in a 3GPP LTE FDD Field Trial

This article presents a summary of 3GPP Long Term Evolution (LTE) radio system performance during the execution of a multi-site field trial. This exhaustive testing campaign, performed over the lower 700 MHz and 2 GHz bands, includes a battery of stationary and mobility throughput cases, multi-user capacity scenarios, handover performance and latency tests, that have been designed to reveal LTE radio performance under a variety of realistic user scenarios. Tests have been performed using a Category 2 LTE UE, which can handle a maximum transport of 50 Mbps in the downlink and 25 Mbps in the uplink. As the results indicate, the system boasts a peak downlink single-user throughput in excess of 49 Mbps and a peak uplink single-user throughput observed of 19 Mbps, which is quite close to the theoretical uplink limit of 21 Mbps. In multi-user scenarios with four active users, in the best case, we have measured a cell throughput of 47 Mbps in the downlink and 20 Mbps in the uplink, which is close to the theoretical best. Finally, as we show, the system achieves a handover success rate of 100% and a best case average user plane latency of 22ms.

Cross Layer Design for OFDMA Systems Using the Beta-Min-Sum Belief Propagation Algorithm

This paper investigates the problem of resource allocation and scheduling for an orthogonal frequency division multiple access (OFDMA) broadband wireless network We formulate a utility-based cross-layer resource management framework that exploits instantaneous channel state information at the base station in order to schedule traffic during each frame. During each decision epoch, the objective is to find an efficient and fair rate and sub-carrier allocation policy that maximizes the average network utility subject to certain constraints. We develop an equivalent graph theoretical model for the cross-layer assignment problem and show its equivalence to the classical problem of finding the maximum weight matching (MWM) on a bipartite graph. This is a well-studied problem in classical graph theory and several well-known solutions exist. For the optimal assignment, we consider the Hungarian algorithm - which always achieves the maximal network utility - and compare its performance to a sub-optimal modified min-sum belief propagation algorithm. Through simulations, we demonstrate the negligible difference in performance between the two.

Timing recovery based on the PAM representation of CPM

A reduced-complexity decision-directed timing recovery method for continuous phase modulation (CPM) is presented. Using the well-known and popular pulse amplitude modulation (PAM) representation-or Laurent representation- of CPM, we develop two different formulations of a PAM-based timing error detector (TED). We consider the general M-ary multi-h CPM model in our development and use the family of aeronautical telemetry CPMs for our numerical examples. We show by analysis that the two TED formulations have identical ldquoS-curverdquo characteristics, and are free of false lock points, but have slightly different performance when applied to specific CPM examples, particularly in the multi-h case. There is no unanimous winner between the two formulations in all cases; however, for both formulations, we show the TEDs are able to perform close to the theoretical limit given by the modified Cramer-Rao bound. As such, the proposed TEDs provide an important synchronization component for reduced-complexity PAM-based CPM receivers that has heretofore been missing.

A New Finite Series Expansion of Continuous Phase Modulated Waveforms

The Laurent Decomposition expresses any binary single-h continuous phase modulated (CPM) signal as the summation of a finite number of pulse amplitude modulated (PAM) waveforms, and the resulting signal space is so constructed that the waveform can usually be synthesized with a reasonable degree of accuracy by using only the ldquomainrdquo component pulse. This derivation has been very useful for reduced complexity demodulation of binary CPM signals. Subsequent to Laurents work, it was shown that commensurate expressions could be obtained for multilevel and multi-h CPM, but with an exponential increase in the total number of PAM component pulses in the signal representation. In this paper, we show that by expressing a CPM signal in its equivalent binary multi- form, we can derive a generalization of Laurents result for the general class of such waveforms that use noninteger modulation indices. In this new signal representation, there is a clearly identifiable data-dependent ldquomainrdquo expansion pulse during each symbol interval which carries most of the signal energy. As in the Laurent Decomposition, the number of terms in this decomposition is only dependent on the CPM signal memory.

Design and performance of a multiple access CPM-SC-FDMA transmission scheme

This paper presents a novel modulation scheme which combines key characteristics of Single Carrier Frequency Division Multiple Access (SC-FDMA) with Continuous Phase Modulation (CPM) to increase the power amplifier efficiency of SC-FDMA transmissions. CPM-SC-FDMA is developed based upon the observation that the samples from a CPM waveform - which are always constant modulus - may be treated as ‘data symbols’ and then DFT-spread for an SC-FDMA style transmission. The result is a new SC-FDMA waveform which generally retains much of the power efficiency of CPM. In particular, we show that two samples per symbol interval of the underlying CPM waveform are adequate to achieve good bit error rate performance. Also, trellis-terminated CPM-SC-FDMA has a low dynamic range in its envelope fluctuations and there is a gain of 6 dB in the peak-to-average power of 4-ary CPM-SC-FDMA relative to QPSK-SC-FDMA.

A generalized CPM-SC-FDMA transmission scheme suitable for future aeronautical telemetry

In this paper, we propose CPM-SC-FDMA, a novel transmission scheme which combines many of the key characteristics of continuous phase modulation (CPM) with Single Carrier Frequency Division Multiple Access (SC-FDMA) in order to produce a power efficient, robust modulation scheme which is suitable for high-rate multiple-access aeronautical telemetry applications. The basis of our approach is found in the observation that the discrete-time samples from a CPM waveform constitute a constant envelope time domain sequence which can be DFT-pre-coded and subsequently mapped to a set of orthogonal subcarriers for an FDMA-style transmission. By a judicious selection of the subcarrier mapping scheme, we show that this new hybrid waveform retains the high power efficiency of CPM. Because it is based on CPM, it is compatible with existing telemetry signaling formats. In addition, it is characterized by a high spectral efficiency due to the concept of overlapping but mutually orthogonal narrowband subcarriers, thus allowing non-interfering multiple-access.

Utility-based cross layer optimization for OFDMA systems using the β-min-sum belief propagation algorithm

This paper investigates the problem of resource allocation and scheduling for an orthogonal frequency division multiple access (OFDMA) broadband wireless network. We formulate a utility-based cross-layer resource management framework that exploits instantaneous channel state information at the base station in order to schedule traffic during each frame. During each decision epoch, the objective is to find an efficient and fair rate and sub-carrier allocation policy that maximizes the average network utility subject to certain constraints. We develop an equivalent graph theoretical model for the cross-layer assignment problem and show its equivalence to the classical problem of finding the Maximum Weight Matching (MWM) on a bipartite graph. This is a well-studied problem in classical graph theory and several well-known solutions exist. For the optimal assignment, we consider the Hungarian algorithm - which always achieves the maximal network utility — and compare its performance to a sub-optimal modified min-sum belief propagation algorithm. Through simulations, we demonstrate the negligible difference in performance between the two.

Power and Spectrally Efficient Multiple Access Using CPM over SC-FDMA

In this paper, we investigate the power efficiency and bit error rate performance of two spectrally efficient CPM-SCFDMA (Continuous Phase Modulated Single Carrier Frequency Division Multiple Access) waveforms. CPM-SC-FDMA is derived by sampling a Continuous Phase Modulated (CPM) waveform and then DFT-precoding the resulting signal samples for transmission using SC-FDMA. Having originated from a constant envelope CPM waveform, CPM-SC-FDMA can potentially retain much of the power efficiency of CPM-thus resulting in a lower peak-to-average power ratio than conventional SC-FDMA. As we show in this paper, when taking the difference in power amplifier backoff requirements into account, CPM-SC-FDMA can provide an overall gain of up to 4 dB relative to convolutionally encoded QPSK-SC-FDMA over a frequency-selective channel.

DCT-based multicarrier schemes for doubly-selective channels

We compare four different multicarrier schemes in a discrete-time, oversampled domain over doubly-selective channels. We find that all the schemes can be implemented in a reduced complexity way, resorting to suitable fast transforms like DFT, DCT, or DST. For all models we assume either a rectangular base pulse or a properly designed prototype pulse, well localized in both time and frequency domains, and we show that such a technique can be applied to our discrete-time block-based model. We highlight the similarities and the differences between some recently proposed multicarrier modulations based on DCT and DST to mathematical concepts such as the Wilson base, that in turn will be used to develop a novel and effective multicarrier format, ready to be employed in practical communication systems. All analyzed modulations will be compared in terms of spectral efficiency and BER performance in practical implementations.

On Reduced-Complexity Soft-Output Detection of Continuous Phase Modulations

We compare low-complexity schemes for soft-output detection of continuous phase modulations (CPMs). In particular, we address the problem of minimizing the complexity of the front end filters and that of the trellis exploited by the detection algorithm, with the aim of assuring a negligible performance degradation with respect to the optimal full-complexity receiver. We show that the approach providing the simplest front end is that based on the CPM decomposition proposed by Moqvist and Aulin. On the other hand, we prove that the most convenient solution in terms of trellis complexity is provided by the CPM decomposition proposed by Mengali and Morelli, possibly combined with suitable techniques for reduced trellis search.