bigyan90
2011-08-26, 10:15 PM
It's a PhD Thesis regarding Uplink Radio Resource Management for QoS Provisioning in Long Term Evolution With Emphasis on Admission Control and Handover.
The Abstract:
Long Term Evolution (LTE) is a beyond 3G wireless system based on a decentralized architecture which shall support end-to-end Quality of Service (QoS). The radio resource management functionalities in the LTE uplink is based on a dynamically shared channel with fast Link Adaptation (LA) including Adaptive Modulation and Coding (AMC) and Fractional Power Control (FPC), Hybrid Automatic Repeat reQuest (HARQ), Packet Scheduler (PS), Admission Control (AC) and handover.
To provide efficient QoS control, it is necessary that both AC and PS are QoS aware. The AC maintains the QoS of in-progress bearers in a cell by admitting a new bearer only if all the existing and new bearers can be guaranteed their QoS requirements. Additionally, LTE will provide seamless access to voice and multimedia services which is achieved by supporting handover.
The problem of providing seamless access becomes even more important in LTE since it uses hard handover (break-before-make type). This PhD study mainly focuses on AC and handover issues for QoS provisioning in LTE uplink.
In the first part, a novel AC algorithm is proposed for LTE uplink to fulfill the required QoS of new radio bearer and in-progress bearers.
In this study Guaranteed Bit Rate (GBR) is considered as the main QoS parameter. The proposed AC algorithm estimates the required resources for the new and existing bearers to fulfill their required GBR taking into account users respective channel conditions.
The proposed AC algorithm is based on a closed-form estimator derived utilizing the FPC scheme standardized in 3GPP. To evaluate the performance of FPC based AC, a reference AC which does not take channel conditions into account is proposed. Furthermore, a QoS aware PS is proposed and
is combined with the AC algorithm for effective QoS provisioning. The performance is evaluated using a full-blown multi-cell, multi-user, semi-static system level simulator following the 3GPP LTE standard. The results show that the FPC based AC, unlike reference AC, is robust and automatically adjusts to the traffic mixes, cell load, and user channel
conditions. Additionally, the proposed AC and PS framework guarantees the respective GBR requirements of different user classes in a best-effort traffic scenario with mixed GBR settings. Further, this framework is shown to guarantee the QoS of users with a realistic Constant Bit Rate (CBR) streaming traffic and an ON/OFF traffic source using CBR traffic to model the ON periods.
In the second part, performance of an intra-LTE hard handover algorithm is evaluated at user speeds of 3 kmph to 120 kmph. Handover algorithm based on Received Signal Strength (RSS) and Carrier to Interference Ratio (CIR) measurements on downlink reference symbols (pilots) is studied. Additionally, Layer 3 (L3) filtering in linear and log-
arithmic domain is evaluated. A realistic estimate of measurement imperfections due to the limited number of reference symbols is modeled and added to the handover measurements before L3 filtering. RSS on reference symbols is known as Reference Signal Received Power (RSRP) and is standardized as one of the measurements for intra-frequency handover in LTE. This study is evaluated using a detailed multi-cell, multi-user, dynamic system-level simulator i.e., a simulator suitable for mobility studies, following 3GPP LTE recommended assumptions. The results show that the downlink measurement bandwidth of 1.25 MHz will lead to best tradeoff between average number of handovers and average uplink Signal-to-Interference-plus-Noise Ratio (SINR). Moreover, it is shown that for an adaptive choice of L3 filtering period, depending on the user speed, the gain for using larger measurement bandwidth can be made negligible for a small penalty on signal quality.
The Abstract:
Long Term Evolution (LTE) is a beyond 3G wireless system based on a decentralized architecture which shall support end-to-end Quality of Service (QoS). The radio resource management functionalities in the LTE uplink is based on a dynamically shared channel with fast Link Adaptation (LA) including Adaptive Modulation and Coding (AMC) and Fractional Power Control (FPC), Hybrid Automatic Repeat reQuest (HARQ), Packet Scheduler (PS), Admission Control (AC) and handover.
To provide efficient QoS control, it is necessary that both AC and PS are QoS aware. The AC maintains the QoS of in-progress bearers in a cell by admitting a new bearer only if all the existing and new bearers can be guaranteed their QoS requirements. Additionally, LTE will provide seamless access to voice and multimedia services which is achieved by supporting handover.
The problem of providing seamless access becomes even more important in LTE since it uses hard handover (break-before-make type). This PhD study mainly focuses on AC and handover issues for QoS provisioning in LTE uplink.
In the first part, a novel AC algorithm is proposed for LTE uplink to fulfill the required QoS of new radio bearer and in-progress bearers.
In this study Guaranteed Bit Rate (GBR) is considered as the main QoS parameter. The proposed AC algorithm estimates the required resources for the new and existing bearers to fulfill their required GBR taking into account users respective channel conditions.
The proposed AC algorithm is based on a closed-form estimator derived utilizing the FPC scheme standardized in 3GPP. To evaluate the performance of FPC based AC, a reference AC which does not take channel conditions into account is proposed. Furthermore, a QoS aware PS is proposed and
is combined with the AC algorithm for effective QoS provisioning. The performance is evaluated using a full-blown multi-cell, multi-user, semi-static system level simulator following the 3GPP LTE standard. The results show that the FPC based AC, unlike reference AC, is robust and automatically adjusts to the traffic mixes, cell load, and user channel
conditions. Additionally, the proposed AC and PS framework guarantees the respective GBR requirements of different user classes in a best-effort traffic scenario with mixed GBR settings. Further, this framework is shown to guarantee the QoS of users with a realistic Constant Bit Rate (CBR) streaming traffic and an ON/OFF traffic source using CBR traffic to model the ON periods.
In the second part, performance of an intra-LTE hard handover algorithm is evaluated at user speeds of 3 kmph to 120 kmph. Handover algorithm based on Received Signal Strength (RSS) and Carrier to Interference Ratio (CIR) measurements on downlink reference symbols (pilots) is studied. Additionally, Layer 3 (L3) filtering in linear and log-
arithmic domain is evaluated. A realistic estimate of measurement imperfections due to the limited number of reference symbols is modeled and added to the handover measurements before L3 filtering. RSS on reference symbols is known as Reference Signal Received Power (RSRP) and is standardized as one of the measurements for intra-frequency handover in LTE. This study is evaluated using a detailed multi-cell, multi-user, dynamic system-level simulator i.e., a simulator suitable for mobility studies, following 3GPP LTE recommended assumptions. The results show that the downlink measurement bandwidth of 1.25 MHz will lead to best tradeoff between average number of handovers and average uplink Signal-to-Interference-plus-Noise Ratio (SINR). Moreover, it is shown that for an adaptive choice of L3 filtering period, depending on the user speed, the gain for using larger measurement bandwidth can be made negligible for a small penalty on signal quality.