Takao Okamawari

Performance Evaluation of End-to-End Communication Quality of LTE

LTE (Long Term Evolution) is standardized in 3GPP (3rd Generation Partnership Project) to provide higher data throughput as well as lower latency for various IP based services including web browsing, VoIP (Voice over IP), video streaming and so on. In this paper, we conduct the performance evaluation of LTE by indoor and outdoor experiments. We evaluated both TCP throughput and UDP transmission delay. For the TCP throughput evaluation, we observed the TCP- uplink-limit problem in which the uplink performance degradation in low SNR (Signal to Noise Ratio) region affects on the downlink TCP throughput. We also analyzed the transmission delay of UDP packets. The delay distribution is affected by the traffic pattern, i.e. packet size and inter packet gap.

Performance evaluation of TCP and UDP during LTE handover

LTE (Long Term Evolution) has been standardized in 3GPP (3rd Generation Partnership Project) to provide higher data throughput as well as lower latency for various IP based services including web browsing, VoIP (Voice over IP), video streaming and so on. In this paper, we evaluate the performance of TCP and UDP when handover is executed in an LTE system. The performance is degraded due to the interference between source and target eNBs (evolved NodeB). Also the handover interruption time negatively affects overall performance. We measure the TCP throughput and the transmission delay of UDP during the handover process by using our indoor and outdoor LTE testbeds. The results show that better TCP throughput performance is obtained by decreasing the A3-offset and the interruption time caused by a handover event is around 80ms.

Design of Control Architecture for Downlink CoMP Joint Transmission with Inter-eNB Coordination in Next Generation Cellular Systems

CoMP (Coordinated Multiple Point transmission) is being discussed in 3GPP (3rd Generation Partnership Project) to improve the cell edge performance as well as the total cell throughput of mobile systems. In this paper, we focus on the control architecture to realize downlink CoMP JT (joint transmission) with inter eNB (evolved NodeB) coordination. We have designed the control architecture with a distributed approach. Since user data is simultaneously transmitted from multiple eNBs to a UE (User Equipment), the schedulers of the coordinated eNBs need to negotiate on resource allocation for CoMP UEs over a backhaul network. This process will be a heavy burden due to the delay of the backhaul network and the complexity of scheduling algorithm that has to take into account a dynamic multi-UE and multi-cell scenario. We propose a simple and practical resource allocation method by which the scheduler in each eNB can make scheduling calculation independently without any resource contention. We also implement the proposed schemes as a real system for evaluation.

A Proposal on Network Control Architecture for CoMP JT with IP Network between eNBs

CoMP JT (coordinated multi-point joint transmission) is being discussed in 3GPP (3rd Generation Partnership Project). User data is simultaneously transmitted from multiple eNBs (evolved NodeB) to a UE (User Equipment) to improve the cell edge performance. The important point here is that not only the throughput of the CoMP UE but also the total throughput of the system must be improved by initiating CoMP JT. Therefore a novel network control architecture is required to optimize the total system throughput by initiating or terminating CoMP JT. In this paper, we propose a network control architecture for CoMP JT that works over an IP backhaul network between eNBs. To mitigate the transmission delay of IP packets, two levels of time scale are introduced. The radio resource for CoMP JT is allocated every several hundreds of milliseconds. On the other hand, modulation and coding scheme (MCS) for link adaptation is calculated every millisecond. The proposed network control architecture is working without a central control entity. This distributed architecture is suitable for LTE-A (Long Term Evolution Advanced). It is also possible to eliminate inter-cluster interference with this proposed architecture.

Delay Performance Analysis of LTE in Various Traffic Patterns and Radio Propagation Environments

Most real-time multimedia services like VoIP (voice over IP), TV conference and real-time gaming require low end-to-end delay. In the near future, it is expected that more delay sensitive applications such as cloud computing or telepresence will appear. LTE (long term evolution) is able to provide real-time services in mobile environments. LTE is designed to have an end-to-end delay of less than 5ms. The delay performance in LTE may, however, be degraded by the radio propagation environment or traffic pattern. For example, if the quality of radio signal is poor, more retransmission is required between eNB (evolved NodeB) and UE (user equipment) which degrades the delay performance. If packet size is large, the packets may be segmented which also degrades the delay performance. Thus it is very difficult to predict the delay performance of LTE system prior to its installation. This paper analyzes how the delay performance in the layer 2 (L2) process of LTE is affected depending on the traffic pattern and radio propagation environment. We conduct numerical simulations that consider the L2 processes. The numerical simulation correctly estimates the delay pattern even in a realistic fading environment.

Experimental Analysis of TCP and UDP during LTE Handover

Mobile IP services, including web browsing, VoIP (Voice over IP), video streaming and so on, are demanding more and more bandwidth with higher communication quality. LTE (Long Term Evolution) is one of the solutions to provide higher data throughput and lower latency. In this paper, we experimentally analyze the performance of TCP and UDP during LTE handover. The mobile users experience performance degradation due to the interference between source and target eNBs (evolved NodeB). Also the interruption during handover may affect delay sensitive service such as VoIP. We measured the TCP throughput and the delay of UDP during the handover process on our indoor and outdoor LTE testbeds. The results show that better TCP throughput performance is obtained by decreasing the A3 offset value. This also improves the delay performance during a handover.

Field Experiment of Network Control Architecture for CoMP JT in LTE-Advanced over Asynchronous X2 Interface

CoMP (Coordinated Multiple Point transmission and reception) is a key technology in 3GPP LTE-A (3rd generation partnership project long term evolution advanced) which is expected to greatly improve the cell-edge throughput performance by the cooperation of multiple eNBs (evolved NodeBs). We developed a prototype system to realize CoMP JT (joint transmission) using a novel network control architecture. The developed CoMP JT system is designed to work over an asynchronous IP-based X2 interface. Therefore we need a simultaneous transmission scheme from multiple eNBs to a UE (user equipment) using the asynchronous X2 interface. We developed a GPS-based (global positioning system based) scheme for the simultaneous transmission. In addition, we developed several new functions to achieve CoMP JT over an asynchronous X2 interface. As a result, a UE can automatically start or stop CoMP JT and the HARQ retransmission is possible during CoMP JT. We conducted a field trial using the developed CoMP JT system. It is demonstrated that the network control architecture implemented in the CoMP JT system works properly over a real asynchronous IP backhaul network.

A novel streaming method using QoS control function of LTE to prevent video freezing

Nowadays mobile video streaming services become very popular. However, due to the drastic bandwidth fluctuation in a mobile network, the quality of the mobile video streaming degrades significantly. Above all, mobile users are not happy when video image freezes for a few seconds/minutes. Most of current video streaming services are working as an OTT (over the top) service, which means a mobile network is regarded as a dumb pipe. Therefore it is very difficult to handle the fluctuation. We propose a novel video streaming method in which a QoS control function of a mobile network is utilized to prevent the video image from freezing. In the proposed method, an original video content is encoded into a sequence of video frames with multiple layers. There are two kinds of layers, a base layer and an enhancement layer. The base layer is small in size but includes all the necessary information to play the original content. The enhancement layer is large in size and provides better image resolution when decoded with a corresponding base layer. Then the base layer is prioritized by the QoS control function in the mobile network when it is delivered to a mobile terminal. On the other hand, the enhancement layer is delivered in a best effort manner. In this way, the mobile terminal can receive the base layer even when wireless signal is degraded and the video image dose not freeze at all. The proposed method is evaluated experimentally using a prototype system. The result of the evaluation shows that the proposed method works properly as we designed.

A Streaming Method for Efficient Bandwidth Utilization Using QoS Control Function of LTE

Recently, mobile video streaming services become very popular. MPEG-DASH is one of the streaming schemes used for the service. In this method, a mobile terminal tries to download video chunks from a streaming server by estimating the available bandwidth of the underlying network. However, it is very difficult to estimate the available bandwidth because this method is working as an OTT (over the top) service, which means the network is regarded as a dumb pipe. Therefore some of the downloaded video chunks are discarded when the estimation is wrong. Since wireless spectrum is a very precious resource, the wasted chunks (downloaded then discarded) should be avoided. We propose a novel video streaming method in which a QoS (Quality of Service) control function of a mobile network is utilized. In the proposed method, an original video content is encoded into multiple layers. There are two kinds of layers, a base layer and an enhancement layer. The base layer is small in size but includes all the necessary information to keep playing the video content. The enhancement layer is large in size and provides better image quality when decoded with a corresponding base layer. Then the base layer is prioritized by the QoS control function in the mobile network when it is delivered to a mobile terminal. On the other hand, the enhancement layer is delivered in a best effort manner. In this way, even if there were no available bandwidth for both of the layers, the mobile terminal can keep playing the video by using the prioritized base layer. Since the enhancement layer is discarded before the wireless transmission when it is congested, the wireless spectrum is efficiently used.

PTP accuracy measurement comparison between boundary clock and VLAN priority

IEEE1588 (PTP: Precise Time Protocol), which aims for the sub-microsecond range of clock synchronization in a minimal network, is currently focused on highly precise clock synchronization. Boundary Clock (BC) and VLAN Priority are considered to be features that improve PTP accuracy. Although some papers addressed the accuracy, these evaluations were realized using different evaluation metrics and environments. We make them comparable to analyse the contribution of BC and VLAN Priority to precise clock synchronization. Our accuracy evaluation used PTP nodes available on the market and shows the results following the test cases defined in the ITU-T G.8261 Recommendation and IEEE1588 standard. The measurement results show that the accuracy was better than 200 nanoseconds with BC enabled, while it was about one microsecond with BC disabled. We observed that the accuracy was improved to approximately 850 nanoseconds when VLAN Priority was enabled.