Jens Steiner

Open Loop Power Control parameter settings impact on LTE HetNet uplink performance

Open Loop Power Control is an important technique providing adaptation of user transmit power. There are multiple factors like cell size, interference conditions, etc. that determine the optimal settings of power control (PC) parameters. In this paper, the impact of open loop PC parameter settings on the performance of LTE uplink (UL) in a co-channel heterogeneous network (HetNet) scenario with macro- and pico-cells is presented. Further, we study the difference in optimal PC settings for various network deployment scenarios and cell range extension (CRE) offsets in order to determine sensitivity of those parameter settings. It is found that optimal PC parameters for one case can serve as good parameters for other network configurations as well. The conclusions are supported by results of system-level simulations.

Antenna tilt and interaction with open loop power control in homogeneous uplink LTE networks

Antenna tilt is an effective inter-cell interference reduction technique which has been studied a lot previously in the CDMA-based systems. In this paper, we concentrate on the evaluations of the mechanical antenna downtilt in the uplink LTE networks. The SINR performance and the trade-off between the cell coverage and capacity are studied in a homogeneous scenario. Simulation results in Macro 1 show, downtilt angle at 12° to 16° has the optimal performance in terms of SINR and the cell coverage is maximized at downtilt angle 16°. The interaction of the antenna tilt with the open loop power control scheme defined in the 3GPP standard is also studied. Results show that, with certain ¿ value, there is no impact to the choice of optimal P0 value with different antenna downtilt angles. For the ¿ value equal to 0.6 and 1.0 respectively, there are about 75% and 37.5% increases in terms of cell coverage and about 47% and 39% increases in terms of cell capacity.

Optimization of QoS-aware Packet Schedulers in Multi-Service Scenarios over HSDPA

This paper illustrates the optimization of design parameters for quality-of-service (QoS) aware packet schedulers in multi-service environments with WCDMA/HSDPA. The optimization sets the objective of maximizing the cell throughput while meeting certain minimum QoS criteria and focuses on two variants of the proportional fair with barrier function (PF-B) family of schedulers and on the proportional fair with required activity detection. The assessment on the sensitivity of the scheduler parameters to the traffic load and mixture is evaluated in scenarios with both real-time (streaming) and non real-time type of users (web browsing). This study shows that the PF-B schedulers require a per scenario dedicated optimization, while PF-RAD is quite insensitive and robust to support a general traffic mix. Overall, PF-RAD significantly outperforms PF-barrier schedulers in the investigated scenarios, providing up to 100% streaming cell throughput gain.

System Level Analysis of Dynamic User-Centric Scheduling for a Flexible 5G Design

In this paper we present our latest findings on dynamic user-centric scheduling for a flexible 5G radio design, capable of serving users with highly diverse QoS requirements. The benefits of being able to schedule users with different transmission time intervals (TTIs) are demonstrated, in combination with a user-centric multiplexing of control and data channels. The proposed solution overcomes some of the shortcomings of LTE-Advanced in terms of scheduling flexibility and performance. In general it is found that using short TTIs is advantageous at low to medium offered traffic loads for TCP download to faster overcome the slow start phase, while at higher offered traffic loads the best performance is achieved with longer TTIs. Using longer TTI sizes results in less control overhead (from scheduling grants), and therefore higher spectral efficiency. The presented analysis leads to the conclusion that a future 5G design shall include support for dynamic scheduling with different TTI sizes to achieve the best performance.

Automatic methods for HetNet uplink power control optimization under fractional load

In this paper we study uplink open loop power control (OLPC) optimization for co-channel Heterogenous Network (HetNet) scenarios with macro and pico cells under fractional load conditions. Two methods for automatic OLPC optimization are proposed, motivated and compared against manual approach. It is found that the OLPC parameters at the pico-layer can be computed as a function of the OLPC parameters at the macro-layer, taking the load at the different cells into account. It is shown that using such a methods results in good overall system level performance. We contribute with in-depth analysis of dependencies between OLPC parameters, cells load, throughput performance and other metrics used to describe network operation on system level.

Improving Dense Network Performance Through Centralized Scheduling and Interference Coordination

Dense network deployments comprising small cells pose a series of important challenges when it comes to achieving an efficient resource use and curbing intercell interference in the downlink. This paper examines different techniques to treat these problems in a dynamic way, from the network and the receiver sides. As a network coordination scheme, we apply a centralized joint cell association and scheduling mechanism based on dynamic cell switching by which users are not always served by the strongest perceived cell. The method simultaneously results in more balanced loads and increased performance. Interference management at the receiver is achieved through the use of a network-assisted interference cancellation and suppression (NAICS) receiver. In order to further boost the fifth percentile user data rates, the transmission rank at the interferers is selectively reduced by a centralized rank coordination functionality. These mechanisms are evaluated in an LTE-Advanced dense small cell scenario with dynamic traffic. Simulation results illustrate that a combination of the centralized cell association and scheduling scheme and interference cancellation at the receiver can provide fifth percentile data rate gains of up to 80% without a detrimental effect on the median user rates, under the applied assumptions and simulation settings. The gains reach 110% when rank coordination is applied.

Interference Management with Successive Cancellation for Dense Small Cell Networks

Network-Assisted Interference Cancellation and Suppression (NAICS) receivers have appeared as a promising way to curb inter-cell interference in future dense network deployments. This investigation compares the performance of a NAICS receiver with successive interference cancellation capabilities, known as Symbol-Level Interference Cancellation (SLIC), with respect to a baseline Minimum Mean Square Error-Interference Rejection Combining (MMSE-IRC) receiver. The study is carried out on a dense, clusterized small cell network, illustrating to which extent NAICS can overcome the additional interference associated with such deployments. Moreover, we analyse how much the data rates could be potentially improved by estimating the throughput with ideal cancellation of the dominant interferer. The results point to limited gains of up to 12% in the coverage data rates when NAICS is used, and that the instantaneous throughput might increase up to 100% in the most favourable case, falling significantly below the estimated 200% maximum gain with ideal cancellation.

Sensitivity Analysis of Centralized Dynamic Cell Selection

Network-Assisted Interference Cancellation and Suppression (NAICS) receivers have appeared as a promising way to curb inter-cell interference in future dense network deployments. This investigation compares the performance of a NAICS receiver with successive interference cancellation capabilities, known as Symbol-Level Interference Cancellation (SLIC), with respect to a baseline Minimum Mean Square Error-Interference Rejection Combining (MMSE-IRC) receiver. The study is carried out on a dense, clusterized small cell network, illustrating to which extent NAICS can overcome the additional interference associated with such deployments. Moreover, we analyse how much the data rates could be potentially improved by estimating the throughput with ideal cancellation of the dominant interferer. The results point to limited gains of up to 12% in the coverage data rates when NAICS is used, and that the instantaneous throughput might increase up to 100% in the most favourable case, falling significantly below the estimated 200% maximum gain with ideal cancellation.

Achieving Ultra-Reliable Low-Latency Communications: Challenges and Envisioned System Enhancements

URLLC have the potential to enable a new range of applications and services: from wireless control and automation in industrial environments to self-driving vehicles. 5G wireless systems are faced by different challenges for supporting URLLC. Some of the challenges, particularly in the downlink direction, are related to the reliability requirements for both data and control channels, the need for accurate and flexible link adaptation, reducing the processing time of data retransmissions, and the multiplexing of URLLC with other services. This article considers these challenges and proposes state-of-the-art solutions covering different aspects of the radio interface. In addition, system-level simulation results are presented, showing how the proposed techniques can work in harmony in order to fulfill the ambitious latency and reliability requirements of upcoming URLLC applications.

Agile 5G Scheduler for Improved E2E Performance and Flexibility for Different Network Implementations

In this article, we present a holistic overview of the agile multi-user scheduling functionality in 5G. An E2E perspective is given, including the enhanced QoS architecture that comes with 5G, and the large number of scheduling related options from the new access stratum sub-layer, MAC, and PHY layer. A survey of the 5G design agreements from the recently concluded 5G Study in 3GPP is presented, and it is explained how to best utilize all these new degrees of freedom to arrive at an agile scheduling design that offers superior E2E performance for a variety of services with highly diverse QoS requirements. Enhancements to ensure efficient implementation of the 5G scheduler for different network architectures are outlined. Finally, state-of-the-art system level performance results are presented, showing the ability to efficiently multiplex services with highly diverse QoS requirements.