Connecting Sensor Networks with IP using A Configurable Tiny TCP/IP Protocol Stack Guangjie Han

Maode Ma

School of Computer Science Shenyang Institute of Chemical Technology Shenyang, China [email protected]

School of Electrical & Electronic Engineering Nanyang Technological University Singapore city, Singapore [email protected]

Abstract—Wireless Sensor Networks (WSNs) consist of a large number of small, smart and low-cost computing devices. Offering a higher degree of flexibility, WSNs are drawing plenty of attention as methods to realize a ubiquitous society. Several sensor networks which are physically deployed in different places sometimes need to be integrated over IP to provide meaningful services for users. It will be widely used in the human society. However, connecting heterogeneous sensor networks with IP is still an open and challenging problem. Recently, implementing TCP/IP in sensor networks attracts much interest nowadays. In this paper, based on the existing researches, we propose a configurable tiny TCP/IP protocol stack integrated with Session Initiation Protocol (SIP) module for a WSN. Using it, we can customize the different protocol set of tiny TCP/IP stack based on the different resource conditions of sensor nodes. Our research has been mostly conducted both in a simulator environment and on physical sensor nodes. The implementation of the configurable tiny TCP/IP protocol stack allows the users to query, control and monitor in different WSNs. Keywords—Wireless Sensor Networks; tiny TCP/IP stack; SIP; Mobility.

I. INTRODUCTION In a WSN environment, sensor networks cannot be separated from an external network which transfers control, query and monitoring messages [2]. With the rapid development of sensor technology, most of the sensor nodes have much more resources and powerful computing capability than before; some of the nodes are expected to be mobile in a modern WSN. For example, in order to monitor the conditions of patients and doctors in a large hospital, we need to attach the sensor nodes to them. The information of the doctors and patients should be transferred to the monitor center in time, and the monitor center can transparently query their conditions and accurately locate them via Internet. Thus, a WSN in the future has the following new requirements: (1) Support of mobility. No matter where the node is, the users can locate it using the uniform address and execute the operation of query, control and monitoring; (2) Name addressing. When the sensor network has many mobile nodes at the same time, it is very hard for the users to remember and record the conditions of every node. If every node has been assigned a name (such as [email protected]) that is easy to remember, then the users can efficiently manage it; (3) Automatic configuration. It is not feasible to configure manually every node in a WSN. Thus, a

1-4244-0983-7/07/$25.00 ©2007 IEEE

node should support the function of automatic configuration and reduce the participation of the users. In the future, all kinds of sensor networks will be integrated into the existing IP networks to provide transparent service for the users and allow the users to query, control and monitor the sensor nodes via Internet. Up to now, since there is no uniform protocol standard in sensor networks, many complex translation problems among different protocols will surface, which are difficult to solve. Nowadays, TCP/IP has become the de-facto standard protocol suite for wired networking. Using TCP/IP is the best and convenient way to connect TCP/IP networks and sensor networks. But, due to the limitation of resources and processing capability, many people believe it is impossible and impracticable to use the TCP/IP protocol suite in sensor networks. However, it was shown in [1] that it is feasible to implement TCP/IP stack on a 8-bit micro-controller with only a few hundred bytes of RAM. In this paper, we reduce the TCP/IP protocol suite and propose a configurable tiny TCP/IP protocol stack based on its resource conditions. SIP is an application-layer protocol that allows multiple end points to establish media sessions with each other. SIP has some characteristics such as protocol independence, flexible naming, and support of mobility, thus it can inherently provide a good framework for user mobility [3]. In this paper, we reduce SIP as much as possible and integrate it with tiny TCP/IP protocol stack together to meet the new requirements of the nodes in a WSN. The rest of this paper is organized as follows. Section 2 lists existing related work about how to interconnect TCP/IP networks and sensor networks. Section 3 proposes the architecture of the configurable tiny TCP/IP protocol stack. In Section 4, we present the implementation of the configurable tiny TCP/IP protocol stack in details, including the reduction of different protocol layer. Section 5 describes the performance of our tiny TCP/IP protocol stack with related work. Section 6 summarizes this work and presents directions for future research. II.

RELATED WORK

There are three kinds of different ways to interconnect TCP/IP networks with sensor networks: gateway-based approach, middleware-based approach and TCP/IP for sensor networks approach.

ICICS 2007

Each sensor network has its own specialized communication protocols, so it is simple to use the gateway to connect a sensor network with a TCP/IP network. Due to the lack of the uniform protocol standard of sensor networks, there is no universal architecture suitable for all types of different sensor networks [4]. When the users want to visit other types of sensor networks, they still need to redesign a new gateway. Thus, a gateway-based approach only solves the problem of interconnecting certain types of sensor networks with TCP/IP networks. Another approach is using middleware to connect sensor networks to IP based wire/wireless networks. The authors in [5] propose one method to interconnect with sensor networks using the VIP Bridge. In fact, it is difficult to convert most kinds of protocols between sensor networks and TCP/IP networks using the VIP Bridge. In order to support much more sensor networks, the architecture of VIP Bridge will become more complicated than before, and it is rather hard to implement such a complicated VIP Bridge which will easily cause processing bottleneck problem and lead to single point failure, and etc. In the future, a sensor node usually has more resources and more powerful computing capability, which will bring many novel changes to sensor networks. From a long-term point of view, implementing TCP/IP in sensor networks is an ideal way to interconnect TCP/IP networks with sensor networks. It is also the trend to unite all kinds of different sensor networks together. The existing architecture of TCP/IP for sensor networks can not flexibly provide a configuration function based on the different resource conditions of sensor nodes. Another of its drawbacks is not being able to support the mobility and flexible naming of sensor nodes [6, 7]. In order to solve the problems, in this paper, we propose a configurable tiny TCP/IP protocol stack with SIP module for a WSN, which can efficiently meet the new requirements of mobile nodes. The users can also flexibly and efficiently customize a suitable subset protocol module of tiny TCP/IP stack for different kinds of sensor nodes. III.

ARCHITECTURE OF TINY TCP/IP STACK

The traditional TCP/IP protocol suite is not suitable to be used in a WSN environment. Some of the protocol modules have to be removed due to the resource limitation of sensor nodes. Meanwhile, its standards and mechanisms should be kept as much as possible, so only a subset of TCP/IP protocol suite is implemented in sensor nodes, which is called tiny TCP/IP protocol stack. We can run these protocol modules in a customized manner; modules are available as APIs and it is the responsibility of the users to enable a subset of them. Moreover, the system design should allow an easy runtime insertion/ removal of these modules. The architecture of the configurable tiny TCP/IP protocol stack is shown in Figure 1.

TCP/IP network HTTP, SNMP, TFTP, SIP

SIP

TCP,UDP

UDP

IP, ICMP

IP, ICMP

UDP

MAC

MAC IP, ICMP MAC

Mobile Node Fixed Node

Sink node

Fixed Node

Mobile Node

Sink node Mobile Node

Fixed Node Fixed Node

Figure 1. Architecture of tiny TCP/IP protocol stack

As seen in Figure 1, our tiny TCP/IP protocol stack consists of many independent protocol modules and can be configured by the users to meet the requirements of different types of sensor nodes. For example, a fixed sensor node with less resource has a few protocol modules. However, the sink node with more resource consists of more protocol modules than a fixed node and a mobile mode. The user directly accesses the web-pages of the sink node using HTTP module via Internet, downloads the files to the sink node using TFTP module, and manages the sink node using SNMP module. No matter where a mobile node is, it is very convenient for the user to access and control it using SIP module which is based on the configurable tiny TCP/IP protocol stack. The lightweight modular architecture of configurable tiny TCP/IP protocol stack has been designed keeping into mind the resource constraint nature of different nodes, making it an apt solution for the development of a WSN. Finally, using the configurable architecture of tiny TCP/IP stack, it is feasible to implement TCP/IP in sensor nodes in a WSN. IV.

TINY TCP/IP STACK IMPLEMENTATION

A. Hardware Platform a) We evaluate our research on the test-bed of the sensor node, Webit1.0, developed at NEU [8]. The hardware platform of Webit1.0 adopts a 8-bit AT90S8515 running at 8 MHz with 8Kbytes of flash program memory and 512bytes of system RAM, the size of webit1.0 is 5.7cm×3.6cm×2.0cm. The hardware platform of the sink node is Webit2.0 [9], which adopts a 8-bit ATMEGA161 running at 8MHz with 16Kbytes of flash program memory and 1Kbytes of system RAM. And a wireless communication module is implemented in both Webit1.0 and Webit2.0. We implement the tiny TCP/IP stack in the different test-bed of the nodes to testify the function of flexible configuration. Our minimized tiny TCP/IP stack just requires a few hundred bytes of memory. For sensor networks, we adopt a spatial IP address assignment mechanism to solve the problem of address assignment. With the assignment mechanism, each node can easily construct an IP address using the location information of the node. Using the spatial IP address assignment in [6] may cause the problem of more than one node having the same address, however, it is beneficial for WSNs to solve the routing problem.

B. Reduction of Network Layer Most network layer protocols in TCP/IP protocol suite aren't suitable for a WSN. Therefore the authors mainly implement IP and ICMP of the network layer. IP is the workhorse protocol of TCP/IP protocol suite and it provides unreliable and connectionless data transmission. In the reducing process of IP, we reset every segment of IP header and simplify the IP header setting and remove the IP options to meet the requirements of sensor nodes. For example, due to the resources limitation of a node in WSNs, the node can not accomplish some complex tasks such as IP fragment reassembly, so this process is moved to the sink node. Thus, It can greatly reduce the burden of a node, although we increase more burdens to the sink node, it is more effective and security to reassemble them in the sink node. Therefore, when sending the data, the node should set the sign of segment to be 0. IP has the ability to broadcast data packets on the local network. Using this broadcast function, a node can inform its information to other neighboring nodes, thus the routing protocol can be greatly simplified in WSNs. ICMP is the protocol designed for the function of control, test and management. It has many message types, which are determined by the type and the code field in the ICMP packet message. In order to know if the node is reachable, the node should respond to the echo request of Ping, so we implement the function of echo request and reply. The user sends echo request from PING to the node and waits for the reply. For a passive node, it only needs to identify the echo request from other nodes and send the reply. C. Reduction of Transport Layer The transport layer has two protocols: TCP and UDP. Due to the energy limitation of sensor nodes, the header overhead in TCP/IP is very large for small packet, so we compress the header compression to improve the transmission efficiency and save the energy of the sensor nodes [10-12]. In this paper, we only discuss the reduction of TCP because of the simplicity of UDP. In general, the state of TCP connection can be checked according to the state transition diagram. The node always gives response back to the user, and it never actively establishes the TCP connection. Thus, we can greatly reduce TCP on the condition that the node can give the correct response to the user. The reduction of TCP state transition diagram is shown in Figure 2. The node always responds to the connection request from the user during the connection process of TCP. The initial state of TCP is LISTEN when the node has no TCP connection request. When the node receives a TCP request, it performs a passive open, receives a SYN, and sends a SYN with an ACK, and then transits to the SYN_RCVD state. When the user receives the correct ACK, the establishment of the TCP connection is over, and then the state of TCP connection is ESTABLISHED. In fact, we only need to provide the LISTEN and SYN_RCVD states for the node during the process of TCP connection establishment. Always keeping the node in the LISTEN state can reduce the waiting time for a new connection, thus the connection speed of the reduced TCP can be guaranteed. Although we don't have the CLOSED and SYN_SENT states compared to the traditional TCP state

transition diagram, the LISTEN and SYN_RCVD states are already enough to describe the connection status between the sensor node and the user. LISTEN recv: SYN

send: SYN , ACK

SYN_RECV recv: ACK

ESTABLISH send: FIN simultaneous close

FIN_WAIT_1

recv : FIN send: ACK

CLOSING

recv: ACK recv : FIN , ACK recv : ACK send: send : ACK send:

FIN_WAIT_2

recv : FIN send: ACK active close

TIME_WAIT 2MSL timeout

Figure 2. Reduction of TCP state transition diagram

The termination of a TCP connection is more complicated than the establishment of a TCP connection. After the node receives the last datagram, it sets the FIN state and terminates the current connection. At this time two possible situations could happen: active or passive termination of the connection. Thus the corresponding states to active and passive termination should be considered for a node. Maintaining the TCP state transition diagram is very important to the implementation of much more complicated TCP. The main method is to keep the necessary states and discard the unwanted ones of TCP state transition diagram in accord with the actual characteristic of the node and to implement the transition of TCP state by estimating the possible situations. In the process of reducing TCP, we also keep the simple mechanism such as congestion control and flow control to guarantee the stability of transmission [8, 9, 13]. D. Reduction of Application Layer The application layer has many protocols in TCP/IP protocol suite. We studied on the implementation of HTTP, SNMP, and TFTP in tiny TCP/IP stack. In this paper, we will only discuss the reduction of HTTP. HTTP is a standard Internet protocol, which is used to transfer files. One HTTP supports two types of requests: complete request or simple request. The simple request is enough to monitor the nodes and access them. We choose the "GET" method to implement the request-reply mechanism. How to save memory space as much as possible? A passive sensor node doesn't care about the redundant information except the "GET" line. When the user accesses the sensor node using the browser, it will submit the request without the file name: GET/HTTP/1.0. Most nodes will interpret this request as the one for the default file. So, omitting the parameters of the GET method doesn't significantly affect the response of the HTTP request. E. Reduction of SIP A SIP infrastructure consists of user agents and network servers. A user agent is an entity that acts on behalf of someone who wants to participate in message exchanges. A SIP user

agent is identified by its SIP URL (such as uer@host or user@domain), which is essentially a combination of attributes. The URL can refer to an individual entity or a group of endpoints. In addition to user agents, there are two kinds of network servers: proxy server and redirect server [13]. SIP is independent of low layer protocol and can support TCP or UDP. We reduce SIP as much as possible, while keeping some fundamental methods such as REGISTER, INVITE, ACK and DO and remove unnecessary parts of SIP and guarantee to meet the requirements of mobile nodes. In order to save memory space as much as possible, the reduced SIP runs directly on the top of UDP. We also simplify the SIP architecture and integrate the function of the redirect server into the monitor server, and choose the sink node as the proxy sever of sensor nodes in the domain. When the registration server gets the new registration message form a mobile node, it will transfer the information of the mobile node to the former proxy and the monitor server. No matter where the mobile node is, the user can query, control and monitor the mobile node using the newest registration information. V.

PERFORMANCE

The performance of tiny TCP/IP protocol stack is dependent on a number of variables: hardware and operating environment, application, network bandwidth and traffic load. The performance indices of Webit are code size, run-time memory, the CPU usage, throughput and maximum user connectivity. By evaluating the system performance of Webit1.0, we have the following conclusion: the code size is approximated 3.5 Kbytes, the average run-time memory is 4.8 Kbytes, the CPU usage is about 15%. And by evaluating the system performance of Webit2.0, we have the following conclusion: the code size is approximated 11.6 Kbytes, the average run-time memory is 11 Kbytes, the CPU usage is about 23%. In the sequel, we will compare the features of some similar products to those of Webit. As shown in Table 1, N represents features not supported or Y represents features supported and O represents that we do not find appropriate information about them from the available literature.

respectively. We tested them through applying it to management of a mobile node roaming between two different domains. The result of test is shown in Table II. (Because we can’t get the other similar experimental products, we only tested and compared our products.) TABLE II.

COMPARISON OF TEST RESULTS

Products

Request per second (GETs/s)

Throughput (Kbytes/s)

Processing delay (ms)

Webit 1.0

31

11

43

Webit 2.0

48

20

29

The test results in Table 2 show that our scheme can efficiently meet the data processing requirement of the mobile nodes in a WSN. VI.

CONCLUSIONS AND FUTURE WORK

The authors make the necessary reduction of traditional TCP/IP protocol suite and propose a configurable tiny TCP/IP protocol stack with SIP module for a WSN. Based on the resource conditions of sensor nodes, the users can customize the different protocol module set of tiny TCP/IP stack for different nodes. The technology of the configurable tiny TCP/IP protocol stack provides necessary communication platform for sensor nodes that will be queried, controlled and monitored. Finally, by using the tiny TCP/IP stack for sensor networks, connecting the sensor network with the TCP/IP network is simple and efficient. However, attaching sensor networks to TCP/IP networks may not always be ideal, and a combination of the approach of TCP/IP for sensor networks and middleware, may be beneficial. It not only mitigates the processing burdens of middleware, but also effectively balances the architecture of interconnection networks to avoid the effect of "Large-middleware/Tiny-node". There are still many problems to implement TCP/IP in sensor networks, and we have much more challenging work to do. In the near future, we aim to further develop and optimize the performance of tiny TCP/IP stack using both simulation and experiment with our Weibts. REFERENCES

TABLE I.

Products

uIP [1]

Mobility

COMPARISON OF SIMILAR PRODUCTS Code Size (Kbytes)

[1]

Platform CPU

RAM (Kbytes)

ROM (Kbytes)

AVR

O

O

N

5.2

lwIP [1]

N

21.8

AVR

O

O

[6]

N

O

MSP430

2

60

[7]

N

O

Atemga128

4

16

Webit 1.0

Y

3.5

At90s8515

0.5

8

Webit 2.0

Y

11.6

Atemga161

1

16

As shown in Table I, we can see most products use 8-bit hardware platform, however only our Webit support mobility mechanism. Our Webit1.0 and Webit2.0 only use 3.5 Kbytes and 11.6Kbytes code to accomplish the same function

[2]

[3]

[4]

[5]

A. Dunkels: Full TCP/IP for 8-bit architectures, in Proceedings of The First International Conference on Mobile Systems Applications and Services (MobiSys-03), San Francisco, CA, USA, 2003, 85-98. Mohammad Ilyas and Imad Mahgoub: Handbook of Sensor Networks: Compact Wireless and Wired Sensing Systems, CRC Press, New York, Washington, D.C., 2005. S. Krishnamurthy, TinySIP: Providing Seamless Access to Sensor-based Services, in Proceeding of the 3rd Annual International Conference on Mobile and Ubiquitous systems: networks and services (Mobiquitous06), International Workshop on Advances in Sensor Networks, 2006. A. Dunkels, T. Voigt, J. Alonso, H. Ritter, and J. Schiller: Connecting Wireless Sensornets with TCP/IP Networks, in Proceeding of the 2nd International Conference on Wired/Wireless Internet Communications, February, 2004. Shu Lei, Wu Xiaoling, Xu Hui, Yang Jie, Jinsung Cho, and Sungyoung Lee: Connecting Heterogeneous Sensor Networks with IP based Wire/Wireless Networks, in the 4th IEEE Workshop on SEUSWCCIA'06, 2006, 2561-2566.

[6]

[7]

[8]

[9]

Dunkels A, Voigt T, Alonso J: Making TCP/IP viable for wireless sensor networks, in Proceedings of the European Workshop on Wireless Sensor Networks (EWSN'04), Germany, 2004 Xiaohua Guo, Kougen Zheng, Yunhe Pan, and Zhaohui Wu: A TCP/IP implementation for wireless sensor networks, in Proceedings of IEEE International Conference on Systems, Man and Cybermetrics, 2004 6081-6086. Guangjie Han, Hai Zhao, Jin-dong Wang, Tao Lin, and Jiyong Wang: Webit: A Minimum and Efficient Internet Server for Non-PC Devices, in Proceedings of IEEE Global Telecommunications Conference, 2003 2928–2931. Guangjie Han, Zhao Hai, Wang Jin-dong, and Guan Mo: Study and realization of Webit 2.0 architecture under the Embedded Internet

[10] [11] [12]

[13]

environment, Journal of China Institute of Communication., Vol.25, 2004, 62-68. S. Casner and V. Jacobson: Compressing IP/UDP/ RTP headers for lowspeed serial links, RFC 2508, Internet Engineering Task Force, 1999. V. Jacobson: Compressing TCP/IP headers for low-speed serial links, RFC 1144, Internet Engineering Task Force, 1990. S. Mishra R. Sridharan, R. Sridhar: A robust header compression technique for wireless ad hoc networks, in Proceedings of the 7th ACM international symposium on Mobile ad-hoc networking and computing (MobiHoc'03), 2003. W. Richard Stevens: TCP/IP Illustrated Volume І, II, and III, Beijing: China Machine Press, 2002.

Connecting Sensor Networks with IP using A ...

feasible to implement TCP/IP stack on a 8-bit micro-controller with only a few hundred bytes of RAM. In this paper, we reduce the TCP/IP protocol suite and propose a configurable ... also the trend to unite all kinds of different sensor networks together. .... more burdens to the sink node, it is more effective and security.

229KB Sizes 7 Downloads 239 Views

Recommend Documents

Connecting Sensor Networks with TCP/IP Network
works with an external network by using Application-level Gateways [1] as the inter face. Different .... TCP/IP network to provide meaningful services. In this paper ...

Connecting Sensor Networks with TCP/IP Network
mapping the node label (ID) or location address with IP address in gateway. The IP ... All these information are stored in a database which physically locat-.

Connecting Heterogeneous Sensor Networks with IP ...
into one virtual sensor networks to provide meaningful services .... sensor networks answer “Give me the data that .... save the storage space of the database. 6.

Key Management in IP-based Ubiquitous Sensor Networks - CiteSeerX
corresponding elliptic curve Diffie-Hellman and Digital Signature Algorithm. 6.3.1 Elliptic ... public key signature and verification respectively. This gives us a ...

Key Management in IP-based Ubiquitous Sensor Networks - CiteSeerX
For example, one laptop can easily disrupt the communication of several sensor nodes by ... the sensors, and the malicious node can take control over them [10].

Investigating Sensor Networks with Concurrent ... - IEEE Xplore
The background behind this demonstration is described as an one-page poster submission. The goal is to show a flow of tools for quick sensor network modeling, from an high level abstraction down to a system validation, including random network genera

Fall Detection Using RF Sensor Networks - Sensing and Processing ...
Computer Engineering ... a year. Moreover, falls are the leading cause of injury-related deaths among people 65 and older [1], [2]. ... One type of RF-DFL algorithm is called radio tomographic .... that the person is standing with a high degree of ce

Energy-Efficient Surveillance System Using Wireless Sensor Networks
One of the key advantages of wireless sensor networks (WSN) is their ability to bridge .... higher-level services, such as aggregation and power manage- ment.

Efficient Location Tracking Using Sensor Networks - University of ...
drain-and-balance (DAB), for building efficient tracking hierar- chies, computed from ... and Networking Conference (WCNC) ... Let us call this the hierarchy tree.

On Renewable Sensor Networks with Wireless Energy ...
energy could be transferred efficiently from one source coil to one receiver coil via .... and energy dynamics at a sensor node (Section II-D). B. WCV Traveling ...

Connectivity of Wireless Sensor Networks with ...
Basically, shadowing and fading have so heavy impact on links that there are ... ratio (SNR) γ(x, x ) at the receiver is above a minimal value γmin. In our study, an ...

Robust Location Detection with Sensor Networks
f d. (d) Position of the sensors on the graph. Fig. 1. Our proposed location detection system. ..... Five laptop computers running Red Hat Linux 8.0 are used.

Statistical Location Detection With Sensor Networks
The reader can visualize it as a curve connecting and parameterized by ...... from the fact that the graphs are random at each point, and that. CPLEX uses various ...

Download Building Wireless Sensor Networks: with ...
with ZigBee, XBee, Arduino, and Processing Full. eBook ... data / command mode which can be used to speed up data operations and touches on the use of Xbee ... Programming the Raspberry Pi, Second Edition: Getting Started with Python.

Deploying the BIG-IP System with Microsoft SharePoint - F5 Networks
F5 Analytics (also known as Application Visibility and Reporting or AVR) is a module on the ...... first install and configure the necessary server software for these.

Deploying the BIG-IP LTM with Citrix XenDesktop - F5 Networks
To import the script using Linux/Unix/MacOS systems. 1. Download the script: http://devcentral.f5.com/wiki/default.aspx/tmsh/CitrixXenDesktopMonitor.html. 2.

Deploying the BIG-IP LTM with Oracle Enterprise ... - F5 Networks
May 1, 2012 - http://www.oracle.com/us/products/enterprise-manager/index.html ... 2. Prerequisites and configuration notes. The following are general ...

Deploying the BIG-IP System with Microsoft SharePoint - F5 Networks
What type of network connects servers to the BIG-IP system? (on page 13) ..... 1. On the Main tab, expand iApp and then click Application Services. 2. From the list ...

Deploying the BIG-IP LTM with Microsoft Remote ... - F5 Networks
Aug 16, 2013 - Visit the Microsoft page of F5's online developer community, .... selecting applications that have been published on that page, users initiate new ...... Any other products, services, or company names referenced herein may be ...

Deploying the BIG-IP LTM with CA SiteMinder - F5 Networks
Sep 11, 2012 - proactive health monitoring is critical to the success of all SiteMinder .... 2 You must select Advanced from the Configuration list for this option to ...

Deploying the BIG-IP System v11 with Microsoft ... - F5 Networks
Aug 2, 2013 - See iPhones and other iOS devices are displaying invalid certificate messages after deploying the iApp for ActiveSync on page 58 for important ...

Sensor Data Cryptography in Wireless Sensor Networks - IEEE Xplore
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, VOL. 3, NO. 2, JUNE 2008. 273. Sensor Data Cryptography in. Wireless Sensor ...