Wansu Lim

Dynamic Subcarrier Allocation for 100 Gbps, 40 km OFDMA-PONs With SLA and CoS

The quality of service of 100 Gbps orthogonal frequency division multiple access passive optical networks (OFDMA-PONs) performing dynamic bandwidth allocation is evaluated. New medium access control protocols and frame formats have been developed, exhibiting hybrid OFDMA/time division multiple access scheduling, for capacity enhancement and granular bandwidth allocation. The sequential dynamic subcarrier allocation algorithms allow the network optical line terminal to grant the optical network units (ONUs) bandwidth using both status and non-status based algorithm. Simulations of a 100 Gbps network with 256 ONUs, 256 subcarriers and 40 km extended-reach demonstrate best network throughputs of 87.5 Gbps and 3 ms packet delays for high priority service classes, even at maximum ONU load. In addition, high service level agreement (SLA) ONUs exhibit 1.56 Gbps maximum capacity and 48.82 kbps granularity.

QoS mapping for LTE backhauling over OFDMA-PONs

This paper reports on the quality of service (QoS) mapping mechanisms for converged OFDMA-PON and LTE networks to satisfy the stringent QoS requirements. Various options for exhibiting QoS mapping between the LTE bearers and optical OFDMA-PON queues have been presented. These include, wireless classes forwarded directly to one priority optical queue (so called non-mapping mechanism), 1:1 mapping where class identifier, QoS channel indicator (QCI), levels are mapped to their respective priority optical queues and finally group mapping taking into account bandwidth availability at both enhanced node Bs (eNBs) and optical network units (ONUs) to accommodate optical and wireless users requests. Performance figures, based on system level simulation, are drawn with respect to the QCI delay between eNBs and evolved packet core (EPC) for various numbers of users per cell.

Dynamic Bandwidth Allocation in Heterogeneous OFDMA-PONs Featuring Intelligent LTE-A Traffic Queuing

A heterogeneous, optical/wireless dynamic bandwidth allocation framework is presented, exhibiting intelligent traffic queuing for practically controlling the quality-of-service (QoS) of mobile traffic, backhauled via orthogonal frequency division multiple access-PON (OFDMA-PON) networks. A converged data link layer is presented between long term evolution-advanced (LTE-A) and next-generation passive optical network (NGPON) topologies, extending beyond NGPON2. This is achieved by incorporating in a new protocol design, consistent mapping of LTE-A QCIs and OFDMA-PON queues. Novel inter-ONU algorithms have been developed, based on the distribution of weights to allocate subcarriers to both enhanced node B/optical network units (eNB/ONUs) and residential ONUs, sharing the same infrastructure. A weighted, intra-ONU scheduling mechanism is also introduced to control further the QoS across the network load. The inter and intra-ONU algorithms are both dynamic and adaptive, providing customized solutions to bandwidth allocation for different priority queues at different network traffic loads exhibiting practical fairness in bandwidth distribution. Therefore, middle and low priority packets are not unjustifiably deprived in favor of high priority packets at low network traffic loads. In addition the protocol adaptability allows the high priority queues to automatically over perform when the traffic load has increased and the available bandwidth needs to be rationally redistributed. Computer simulations have confirmed that following the application of adaptive weights the fairness index of the new scheme (representing the achieved throughput for each queue), has improved across the traffic load to above 0.9. Packet delay reduction of more than 40 ms has been recorded as a result for the low priority queues, while high priorities still achieve sufficiently low packet delays in the range of 20 to 30 ms.

Next generation PONs with wireless backhauling

This paper is tackling the issue that stem from the coexistence of diverse technologies, such as emerging OFDMA-PON and WiMAX/LTE, within the same infrastructure. It outlines the key requirements that need to be fulfilled for the backhauling links supporting next generation wireless standards. These links should provide long reach as well as low-cost remote radio base stations coupled with centralised processing. To that extent, the industrial standard simulation platforms were used to evaluate the key network features from the PHY perspective gained by the convergence. The network modelling results confirm the transmission of 16 CPRI signals up to 100 km OFDMA-PON infrastructures achieving 40 Gbit/s total aggregate data rates.

QoS aware MAC protocol for OFDMA-PON

A quality of service (QoS) aware medium access control (MAC) protocol is presented for next generation OFDMA-PONs. The end-to-end delay and network throughput are investigated in the presence of class-of-service and service-level differentiation. In addition, this paper proposes a new dynamic subcarrier allocation (DSA) algorithm. The monitoring window time technique is used where OLT measures the state of the ONUs queue instead of exchanging traditional report messages. As a result, the end-to-end packet delay of high and middle SLAs is less than 0.7 ms even if the ONU offered load is 1.0. The packet delay of the high priority traffic class regardless of SLA levels is however less than 0.6 ms. The total network throughput is 97% of total capacity.

Optical and Wireless Convergence

This paper presents the requirements and possible solutions for wireless convergence in next generation PONs. System level simulation of LTE was performed to evaluate two different approaches for connecting eNBs to ONUs.

Dynamic Subcarrier Allocation for OFDMA-PONs with Monitoring Mechanism

A new protocol design for 10G OFDMA-PONs is reported, demonstrating dynamic subcarrier allocation based on monitoring of each ONUs queuing status. 0.7 ms packet delay and 540 Mbps throughput were achieved for SLA0 ONUs.

MAC protocol requirements for OFDMA-PONs

This paper provides specifications of the parameters defining the design of medium access control (MAC) protocols for orthogonal frequency division multiple access (OFDMA) passive optical networks (PONs). A dynamic bandwidth allocation (DBA) algorithm exhibiting service level agreement (SLA) has displayed packet delays of less than 1.5 ms for the middle and high SLA optical network units (ONUs) at 156 Mbps, representing the maximum ONU offered load. A spectral efficiency of 93.75% the total network capacity of 40 Gbps has been achieved, for 256 ONUs and long-reach link lengths of 100 km. The new protocol functionalities have been attained by the extension of the XG-PON frame structure to demonstrate OFDMA/TDMA granularity.

Motor-imagery EEG signal classification using position matching and vector quantisation

This paper proposes a motor-imagery electroencephalograph (EEG) signal classification method using vector quantisation and position matching. An EEG signal is transformed into a sequence of feature vectors which are extracted from the signal by short-time Fourier transform (STFT). A sequence of feature vectors, in turn, is transformed into code index sequence using the code book which is created by vector quantisation through Linde-Buzo-Gray (LBG) algorithm. Test signals are classified by K-nearest neighbour (KNN) where similarity among vectors is obtained by position matching. Brain computer interface (BCI) competition 2003 dataset III was used in the proposed method to perform the classification test and the maximum performance of 83.57% was obtained.

Noise reduction techniques using particle filters in brain computer interface systems

This paper describes a particle filter design to reduce non-Gaussian noise for classification improvement in brain computer interface (BCI) systems. The particle filter is a type of non-parametric filter that can find samples repeatedly if there is nonlinear and inaccurate information in the system. In order to verify the performance improvement using the particle filter, we simulated data in a nonlinear BCI system. The algorithms of the common spatial filtering (CSP) for maximising the class variance differences, linear discriminant analysis (LDA) for reducing the dimension number in data analysis, and support vector machine (SVM) for statistical data classification using hyperlanes were used. When using the particle filter in the motor imagery class from the BCI system, the data was more accurately categorised in the data class. Therefore, we confirm the significant classification improvement from the line of the demarcation and focal planes when using the particle filter in the motor imagery class from the BCI system.