Myungmoon Lee

Design of hierarchical crossconnect WDM networks employing a two-stage multiplexing scheme of waveband and wavelength

The authors propose a design method for hierarchical crossconnect wavelength-division-multiplexed networks employing a two-stage multiplexing scheme of waveband and wavelength, in order to reduce the complexity and size of optical crossconnect at the nodes. The waveband is formed by grouping lightpaths with the same destination in a network. The authors present an integer linear programming formulation for the routing and grouping of lightpaths, and a heuristic design method is also proposed as a practical method to find a design solution for general large-scale networks. It is found that the proposed design method can largely reduce crossconnect requirements, although it slightly increases the wavelength requirements.

Multiple silicon nanowire complementary tunnel transistors for ultralow-power flexible logic applications

Based on experimental and simulation studies to gain insight into the suppression of ambipolar conduction in two distinct tunnel field-effect transistor (TFET) devices (that is, an asymmetric source-drain doping or a properly designed gate underlap), here we report on the fabrication and electrical/mechanical characterization of a flexible complementary TFET (c-TFET) inverter on a plastic substrate using multiple silicon nanowires (SiNWs) as the channel material. The static voltage transfer characteristic of the SiNW c-TFET inverter exhibits a full output voltage swing between 0 V and Vdd with a high voltage gain of ∼29 and a sharp transition of 0.28 V at Vdd = 3 V. A leakage power consumption of the SiNW c-TFET inverter in the standby state is as low as 17.1 pW for Vdd = 3 V. Moreover, its mechanical bendability indicates that it has good fatigue properties, providing an important step towards the realization of ultralow-power flexible logic circuits.

WDM/SCM MAC protocol suitable for passive double star optical networks

A new WDM/SCM access protocol with node-grouping scheme suitable for passive double star packet networks is proposed. Combining SCM with WDM enables us to place control and data channels on the same wavelength resulting in large bandwidth and high data rate transmission. In the proposed protocol, packet delay is reduced in a large scale because the parts of access control should be done for the number of groups rather than the number of network nodes. Furthermore, additional reduction of the packet delay is achieved by limiting the propagation paths of control packets to the roundtrip between source nodes and sub-hub based on double star network configuration.

Cost-effective design of GMPLS networks with sparse multi-granularity optical cross-connect

In this paper, we propose the efficient placement of MG-OXC (multi-granularity optical cross-connect) nodes in GMPLS networks to reduce all the network cost. When only a limited number of nodes are allowed to have the capability of waveband switching, it is shown that selection based on max-traffic performs better than Random and Nodal-degree schemes. Performance of the sparse placement of max-traffic scheme can be achieved very close to that of all placement of MG-OXC.

A restoration method independent of failure location in all-optical networks

In this paper, a simple and fast restoration method independent of failure location in all-optical networks is proposed. The proposed method uses a single fixed restoration wavelength for any link failure and employs edge-disjoint path restoration (DPR). The restoration process is triggered just after failure detection without the need to identify failure location and accelerated by a parallel crossconnection algorithm. Since fixed wavelength transmitters can only be placed at nodes, the node cost is less than counterparts with tunable wavelength transmitters. Also, transmission cost is only slightly increased because the wavelength requirements in the proposed method are similar to the ones using tunable restoration wavelengths.

Design and evaluation of survivable WDM networks by joint optimization method

Routing and planning of working capacity and spare capacity in WDM networks were investigated. Joint optimization technique has been used and the problem has been formulated to restore the general failure scenarios. The difference between the cost for the restoration of one node failure and the cost for the restoration of one link failure is decreased as more wavelengths per fiber are used. The difference between the cost of virtual wavelength path (VWP) network and the cost of wavelength path (WP) network is increased as more wavelengths per fiber are used.

Analysis of electrical parameters of p-channel silicon nanowire transistors with selectively thinned channels on plastics

We present a technique for thinning the channel region of silicon nanowires (SiNWs) selectively while maintaining a thickness of the source/drain (S/D) regions in an attempt to minimize the parasitic series resistance of SiNW transistors (SNWTs). By transferring the as-fabricated SiNWs onto a plastic substrate, p-SNWTs were fabricated on a plastic substrate, and carrier transport in p-SNWTs was investigated by extracting electrical parameters using the YΦ method, which include mobility attenuation factors, parasitic series resistance (Rsd), and effective channel resistance. It is shown that, in the strong inversion region, the parameters fit the measurement data well and that degradation in device performance in our p-SNWTs under high transverse electric fields is dominated by surface roughness scattering, with minimal Rsd impact on it due to the relatively thick S/D regions.

WDM network design with waveband and wavelength multiplexing scheme

A network design method employing a two-stage multiplexing scheme which exploits the merits of wavelength granularity is presented for wavelength division multiplexing mesh networks. The effects of wavelength granularity and average demand on the crossconnect requirements and wavelength requirements are discussed.

WDM/SCM multiple access protocol with high throughput and low packet delay for passive double star networks

This paper investigates a hybrid wavelength division multiplexing (WDM) multiple access protocol combining the advantages of preallocation and reservation-based schemes using a subcarrier multiplexing (SCM) technique for passive double star optical networks. We include a node-grouping scheme in the proposed protocol to increase channel utilization and to reduce sensitivity to queueing delay time and propagation delay time. All nodes in a network are connected through passive star sub-hub forming multiple groups, and the groups are then connected through a central passive star hub resulting in a passive double star network. Each group is pre-assigned a WDM channel as its home channel, and a specific SCM channel identifies each node. In the proposed protocol, the frame is divided into slots corresponding to the number of groups, not the number of nodes. The reduced number of slots results in reduction of queueing delay time. Furthermore, additional reduction in the propagation delay time of the control packets is achieved by limiting the propagation paths of the control packets to the roundtrip between source nodes and sub-hub in a double star network.

Molecular characterization of a cDNA encoding disulfide isomerase during barley kernel development

We conducted differential hybridization using two different tissues (caryopses as a tester and pericarps as a driver) 14 days after fertilization (DAF) to characterize the molecular mechanisms inherent to barley kernel development. Genes that were predominantly expressed in caryopsis were then sequenced and divided into nine categories based on their putative functions. The transcripts of Hordeum vulgare disulfide isomerase (HvPDI), a cDNA encoding disulfide isomerase, were detected in abundance in the caryopses but were expressed only weakly in the pericarps, stems, and leaves. Additionally, expression of HvPDI was abundant in 5-DAF kernels; however, its expression gradually decreased to 20 DAF. In situ hybridization revealed that HvPDI transcripts were detected primarily in the starch endosperm, located near the aleurone layer, during the later stages of kernel development (i.e., 14 and 20 DAF). The HvPDI gene was up-regulated as a result of several different plant hormone treatments, including treatment with ABA, GA3, benzyladenine, and methyl jasmonate. These findings provide insight into the molecular mechanisms inherent in kernel development and assembly of seed storage proteins for accumulation into caryopses.