In this paper, a Software Defined Networking (SDN) based unified wireless network architecture is proposed to enable distributed mobility management. The proposed approach envisions a virtualized control plane, placed in the Cloud. The user plane is free of typical tunnelling mechanisms and utilizes layer 2 and layer 3 traffic forwarding. The network architecture for cellular users is described first. Then, traffic flow in both upstream and downstream directions, different mobility scenarios, and an efficient address assignment scheme are discussed. An analytical model to estimate the signaling overhead in the Evolved Packet Core (EPC) is presented. The architecture is further extended to support WLAN users in fixed networks, with a description of the signaling schemes required in a number of mobility events.

This paper presents a software defined networking (SDN) based semi-distributed mobility management system. By both using an SDN approach and leveraging layer 2 technology, i.e., carrier grade Ethernet and the 802.1ad standard, it is possible to simplify the network architecture by removing the need for the relatively complex GPRS tunneling protocol currently in use in the user plane. An analytical model is used to evaluate the control overhead of the proposed approach and a proof of concept testbed based on OpenFlow running the Floodlight controller is presented. Obtained results demonstrate that, as long as the SDN controller can quickly react to user mobility, frame loss can be mitigated and the proposed architecture is viable.

In cellular networks it is the mobile node's responsibility to update the network about its location change, especially when this one enters idle mode. We developed a new framework [8] where the idle interface is powered-off to save energy and thus could neither detect its location change nor perform a location update. This framework uses a proxy entity at the network and at the mobile node the active radio interface for proxying the idle interface. In this paper, we present an approach that relies on the proxy entity and the active interface for proxying the location update procedure of the proxied interfaces. More specifically, we propose two algorithms. The first algorithm considers proxying periodic location update of idle interface without considering its mobility if there is any. The second algorithm is based on sending the location to the Information Server which determines the list of paging/tracking areas that serves the current location. We present the design of our architecture for 4G systems, namely WiMAX and LTE. The algorithms are analytically evaluated to evaluate the power savings compared with single-radio power management.

The new generation of handheld devices is equipped with multi-radio interfaces that enable radio link connections to a variety of wireless networks (i.e. Wi-Fi, WiMAX).This allows for seamless connectivity; however, this also raises a serious issue concerning the short longevity of the handheld usability due to the high power consumption of the wireless interfaces. In this paper, we present a framework that increases considerably the battery longevity of the handheld devices; it enables efficient power management from a global point of view (and not a single radio interface view) of multi-radio devices. The basic idea behind our proposal is to power off the idle interface but at the same time to keep it in virtual idle mode in the network by extending IEEE 802.21 on both sides (the mobile node side and the network side). On the network side, the 802.21 entity acts as a proxy of the powered-off interface to insure the assigned resources for the interface are always maintained during the proxying period. In the context of the proposed framework, we present (details of) the mechanisms to proxy an idle interface (after powering it off) and to wake up a proxied interface respectively. The proposed solution is analytically evaluated to quantify the power savings compared with single-radio power management.

We present a new task dissemination scheme called multicast deluge (MDeluge). MDeluge can disseminate tasks into a subset of the sensor network. MDeluge uses a tree which is formed when the source node broadcasts the code version message. The source node keeps the tasks and sends it based on the requests. MDeluge supports newly emerging wireless sensor network architectures with layered structure and mobile sensor nodes. We developed a prototype implementation of MDeluge on TelosB motes. Assuming a grid structured wireless sensor network, we present an analysis of various message costs as well as the overall cost of data messages. Experimentation with the prototype and simulation of MDeluge are done to show the MDeluge performance in static sensor network. Simulation also shows how the performance will be impacted by various parameters in a sensor network with mobile sensor nodes.

The paper presents an analysis of the tracking component of SIP paging that has recently been introduced. Tracking protocol enables the tracking agent to keep track of the mobiles’ handoffs between access points. The tracking protocol can be modeled using soft-state and hard-state approaches. We analytically compare each of these approaches for state inconsistency ratio, message rate and the overall cost and determine the ranges of system parameters for which the soft-state approach is superior. Simulation model we developed enables us to evaluate the traffic introduced by the tracking protocol and the cache (state) size.

CAPWAPHP is a new handover and context transfer protocol for WLAN hosts managed by an access controller. The handover protocol is designed to ensure that the station has an association with a single wireless termination point and when the station does a handover to another wireless termination point, ensure that forwarding tables of the switches are updated. The context transfer protocol is designed as an original application of IETF's context transfer protocol using neighbor graphs. Station authentication context which is established at the association time gets transferred to neighboring wireless termination points and as a result, in subsequent reassociations the authentication time can be reduced. For Split MAC wireless termination points the context is stored at the access controller. Soft-state Markov chain analysis results indicate that context hit rates increase with average residence times under a wireless termination point and simulation results indicate that much shorter handover latencies result compared with full authentication without context transfer.

Wireless Electroencephalograms (EEG) are currently being used to wirelessly transmit the data from brain sensors to a computer and they carry huge potential for many future medical applications. This paper presents the design of a hy-brid medical sensor network with Tmote Sky motes as wire-less EEG sensor nodes at the lowest level collecting EEG signals and sending them to Stargate PDAs at the next level. Stargates perform artifact removal, Fourier transfor-mation and feature extraction and the final machine intelli-gence algorithms are run at a PC server. Several features of the CodeBlue medical sensor network like query processing, routing layer are used in our design. The advantages of the hybrid medical sensor network integrating wireless EEGs include the capability to have the brain monitoring func-tionality incorporated into the medical sensor networks. ACM Classification

In today's wireless sensor networks there is a need to improve macro programming of the sensor nodes so that sensor nodes can execute various applications in a flexible manner. There is a need to contribute to macro programming of wireless sensor networks by developing new code dissemination techniques which can disseminate the code into designated sensor nodes. We present a new algorithm called multicast deluge (MDeluge) which can be used to disseminate the code image into a designated subnet of a wireless sensor network. MDeluge disseminates code in the sensor network using a tree which is formed when the sensor nodes send code request messages. Micro server keeps the code and sends it based on the requests. MDeluge disseminates binary code as well as capsules of Mate. We extend MDeluge for the sensor networks that are geographically distributed and that have moving nodes. Assuming a grid structured wireless sensor network we present an analysis of various message costs as well as the overall cost of data messages of MDeluge. Simulation of MDeluge shows that MDeluge performs better than Deluge to disseminate the code into designated sensor nodes

The paper addresses location tracking and paging IEEE 802.11 wireless LAN hosts that are in power save mode. We design a protocol for paging (tracking agent based paging or TAP) IEEE 802.11 wireless LAN hosts. Mobile nodes in each IEEE 802.11 basic service set constitute a layer-2 paging area, which coexists within layer-3 paging areas. The tracking agent is in charge of the L2 paging state, and it continuously updates the paging state after each L2 handoff. Paging on the wireless link is done by extended beacons, and does not involve periodic L3 messages. Signaling in TAP is based on IEEE's inter-access point protocol with minor extensions. We show that TAP can be integrated with various IP paging algorithms such as mobile IPv6 hierarchical paging, home agent and foreign agent paging. We evaluate the performance of the combined L3-L2 paging protocols analytically and by simulation and show that TAP combined with mobile IPv6 hierarchical paging achieves superior power savings compared individually to mobile IPv6 hierarchical paging without TAP or home agent paging algorithms.

We design a methodology for fast IPv4 and IPv6 address acquisition by mobile layer-2 entities in a wireless LAN. The main idea is to do the IP address assignment whenever it is necessary, not after the mobile establishes a layer-2 connection with a new access point. In order to achieve this, we propose to encapsulate the current IP address of the mobile to the association or re-association message in 802.11 WLANs. We present procedures of address acquisition for IPv4, IPv6 and Mobile IPv4, Mobile IPv6. Due to the hidden station problem, the mobile cannot successfully achieve the duplicate address detection (DAD) procedure in IPv6. Our proposal does not require the DAD, so the impact of the hidden station problem in IPv6 address assignment is eliminated. The proposed address assignment procedure, which is executed with layer-2 association, is faster than the existing DHCP-based stateful IP assignment.

We design a protocol for paging (tracking agent based paging or TAP) IEEE 802.11 wireless LAN hosts. Mobile nodes in IEEE 802.11 BSS constitute a layer-2 paging area. The tracking agent (TA) is in charge of the paging state, and TA continuously updates the paging state after each layer-2 handoff. Paging on the wireless link is done by extended beacons, and does not involve periodic layer-3 messages. Signaling in TAP is based on IEEE's IAPP (inter-access point protocol) with minor extensions. We evaluate the performance of TAP and show that TAP achieves superior power savings.

The paper addresses dormant mode operation support for wireless local area network (WLAN)-UMTS roaming for dual-mode users. Dormant mode operation means that the dormant MS stays in its dormant mode when it roams from WLAN to UMTS or UMTS to WLAN and keep its sessions open using mobile IP. Current technology doesn't support dormant mode operation when users roam between inter-technologies. User may roam in one of the technologies only. We propose a methodology to support the dormant mode roaming between inter-technologies, namely wireless local area networks and UMTS (or GPRS). The methodology is based on letting the access point of WLAN or GGSN of UMTS perform necessary registration functions in lieu of the mobile.

A novel application on the recently emerging Personal Communication Services (PCS) networks is the multimedia communication.In this paper; we evaluate multimedia communications capability of one of the PCS standards, Personal Access Communications System's (PACS) packet channel (PPC) using simulation modeling.The performance of PPC's slot aggregation and data-sense multiple access techniques are studied by considering the downlii and uplink in a single cell and combined uplink/ downhnk in two cells and changing various parameters such as the number of users and certain protocol parameters.Interconnection of PPC with the Internet is discussed next.Frame rates of MPEG-1 coded images transmitted in a PACS cell as IP datagrams are determined.Handover characteristics of PPC downlink are studied by changing different parameters such as the cell size, the speed of the mobile host and time between handovers.The results clearly establish that multimedia communication on PPC is only feasible at lower bandwidths and frame rates and only a few users per cell can be supported.Careful tuning of PPC protocol parameters are required.There is one parameter whose variation gives opposite results on the downlink and uplink.

Timed communicating sequential processes (TCSP) language is used to specify fine-grain and coarse-grain multimedia synchronization. A lip synchronization system is an example of fine-grain synchronization. Several groupware scenarios are examples of coarse-grain synchronization. The formal specifications are used as the basis of verification and simulation. Safety and liveness timing requirements of the synchronization system are stated in terms of temporal logic formulas. Correctness analysis of the specification is shown using the temporal formulas and TCSP proof theory. It is shown that TCSP is powerful enough to be used in multimedia system design and verification. Next, simulation of multimedia synchronization is discussed. Various simulation models are developed for fine- and coarse-grain synchronization systems. It is shown that simulation modeling can lead to early detection of possible synchronization violations. The buffering requirements of a given synchronization mechanism can be effectively studied using simulation.

We develop a method for specifying the behavior of network management agents for the Internet framework using SDL-92 which is object-oriented. The starting point is the definition of a management information base (MIB) in ASN.1 with macros as well as the description of the simple network management protocol version 2 (SNMPv2 protocol). The behavior of an agent is defined in an SDL-92 process which dynamically creates processes for the rows of the tables of the MIB and in several procedures that process variable bindings in protocol data units (PDUs) and carry out specific operations on conceptual rows such as creation, retrieval, and bulk retrieval. Some reusable, MIB-independent process types and procedures are identified. For large MIBs, we first obtain a class diagram representation of the MIB. From the class diagram, the classes with behavior are mapped to process types representing conceptual rows of specific tabular objects. SNMPv2 protocol operations on these tabular objects can be easily specified with the help of MIB-independent procedures. Next, considered is the generation of test cases. MIB test cases can be generated from the SDL-92 specification using specification analysis and behavior inversion techniques. Many test cases for an example single-table MIB are designed and specified in tree and tabular combined notation (TTCN) based on the specification in SDL-92.

Encoding/decoding is a functionality which takes about half of the protocol processing. ASN.1 is a standard language to define messages of the application/layer. BER is the standard that describes encoding/ decoding of PDUs defined in ASN.1. The authors show that decoding can be divided into two phases: parsing and type-value matching. Algorithms are developed for these transformations. Distributed implementation of these algorithms can be achieved with specialized encoders/decoders. Four different architectures are defined for single-layer encoding/decoding. One of these architectures defined as VLSI-based ASN.1 encoder/decoder (VASN1) achieves better parallelisation as well as pipelining of parsing with type-alue matching subphases. VASN1 is composed of 16 instruction RISC processors. The processors used as execution units perform subphase transformations while the processor used as central controller assigns workload to the execution units. The processor used as interface unit takes care of input/output.>

Presents test selection criteria for Prolog programs which are based on control flow. The control flow in Prolog programs is not obvious because of the declarative nature of Prolog. The authors present two types of control flow graphs to represent the hidden control flow of Prolog programs explicitly. A fault model is developed for Prolog programs for guidance on test selection. Test selection criteria are given in terms of the coverage on these control flow graphs. Under the given fault model, the effectiveness of these criteria is analyzed in terms of fault detection capability of the test cases produced with these criteria.>

An experience with formal specification, analysis, and testing of an application layer protocol is presented. The ISO file, transfer, access, and management (FTAM) protocol is chosen due to its potential for widespread use. The specification language used was the ISO standard Estelle. This tool generates control and dataflow graphs of the specification and derives unparameterized test sequences for each function identified by the user. The authors describe formal specification of application layer protocols in Estelle and translation of ASN.1 data definitions into Estelle data types. The test design tool is used to obtain functional decomposition of the control and dataflow graphs. This way unparameterized test sequences are obtained. These sequences lead to a complete test suite obtained by parameterization which must be the next step. Analysis of the control and dataflow graphs leads to the derivation of several properties that most of the application layer protocols must possess. The identified properties are shown to simplify the test design process.>

A novel approach is proposed for generating test cases from LOTOS specifications for protocol conformance testing. A set of transformation rules is applied to the specification, and the specification is mapped to an extended finite state machine called a chart. The flow of data in the chart is modeled by a data flow graph. Test cases are generated from the chart. Program testing strategies are applied for test case selection. A tool implementing this methodology on a workstation is also discussed.>

An experience is presented with an automated test design tool for functional analysis and test derivation of distributed systems formally specified using Estelle, a description technique based on an extended finite-state machine model. The tool accepts a formal specification of the system and generates control, data flow graphs, and unparameterised test sequences. The tool has been used, on an experimental basis, for conformance test design of ISO File, Transfer, access and Management protocols.>

A computer-aided design tool is presented for designing test for protocols. CAD-PT accepts a formal specification in Estelle of the protocol and generates control and dataflow graphs on a graphic workstation. The tool is explained component-by-component, using a simplified ISO Class 2 transport protocol as an example.>

Communication protocol testing can be done with a test architecture consisting of remote Lower Tester and local Upper Tester processes. For real protocols, tests can be designed based on the formal specification of the protocol which uses an extended finite state machine model. The specification is transformed into a simpler form consisting of normal form transitions. It can then be modeled by a control and a data flow graph. The graphs are decomposed into subtours and data flow functions, respectively. Tests are designed by considering parameter variations of the input primitives of each data flow function and determining the expected outputs. The methodology gives complete test coverage of all data flow functions and control paths in the specification. Functional fault models are proposed for functions that are not formally specified.

The paper discusses the use of formal specifications for conformance testing of OSI protocols and divides the discussion in two parts: test design and tester design. A draft standard formal specification of the Class 4 transport protocol in Estelle is taken as the starting point for test design. The test design technique used to derive a conformance test suite is semi-formal based in part on the formal specification and also the informal specification. The tests obtained are expressed in natural language. In the tester design part, we discuss the distributed test architecture of ISO and give the detailed designs of upper and lower testers. It is shown that, in testing Class 4 transport implementations, a parametrized protocol implementation approach in the lower tester design, renders the tests easier to implement.

Protocol testing for the purpose of certifying the implementation's adherence to the protocol specification can be done with a test architecture consisting of remote tester and local responder processes generating specific input stimuli, called test sequences, and observing the output produced by the implementation under test. It is possible to adapt test sequence generation techniques for finite state machines, such as transition tour, characterization, and checking sequence methods, to generate test sequences for protocols specified as incomplete finite state machines. For certain test sequences, the tester or responder processes are forced to consider the timing of an interaction in which they have not taken part; these test sequences are called nonsynchronizable. The three test sequence generation algorithms are modified to obtain synchronizable test sequences. The checking of a given protocol for intrinsic synchronization problems is also discussed. Complexities of synchronizable test sequence generation algorithms are given and complete testing of a protocol is shown to be infeasible. To extend the applicability of the characterization and checking sequences, different methods are proposed to enhance the protocol specifications: special test input interactions are defined and a methodology is developed to complete the protocol specifications.

Protocol testing for the purpose of certifying the implementation's adherence to the protocol specification can be done with an architecture which includes a remote Tester and a local Responder processes generating specific input stimuli called test sequences. It is possible to adapt test sequence generation methods of finite state machines, namely transition tour, characterization and checking sequence methods to generate test sequences for protocols specified as incomplete finite state machines. For certain test sequences, the Tester or Responder processes are forced to consider the timing of an interaction in which they have not taken part; these test sequences are called nonsynchronizable. The three test sequence generation algorithms are modified to obtain synchronizable test sequences. Checking the protocol designs for intrinsic synchronization problems is also discussed.

Experience with the use of formal descriptions of communication services and protocols is described. The paper focuses on the experience of the authors with the extended state transition model which is proposed as a standard formal description technique (FDT) for the services and protocols in the OSI environment. The first part of the paper refers to various example specifications, including transport protocol and service specifications, and discusses the suitability of the specification method and possible extensions. In the remaining part, the use of such formal specifications during the phases of system design, implementation, and testing is described. Various approaches to protocol design validation, implementation, and assessment of implementations are discussed, with emphasis on the last point. The experience with several of these approaches is described in the paper, and further details may be found in the references.