Average Reviews:
(More customer reviews)The Mobile in the book's title refers to 2 large scale cases. Unsurprisingly, one case is the many millions of people who have mobile phones and who use these for wireless communication while moving around some region. The other case is a Wireless Sensor Network [WSN], where nodes (aka. sensors) are deployed in the field. Perhaps a little ironically, the nodes need not be mobile. Imagine for example where we have nodes scattered around a wilderness area, to detect animals and bushfires. While the chapters that consider WSNs also cite some instances of actual mobile nodes, most of the examples are static. This discrepancy with the title does not seem to be explicitly addressed anywhere in the text, but you might want to keep it in mind while reading.
Chapter 4 talks about reducing channel underutilisation in a Mobile Ad Hoc Network [MANET]. While interesting, on page 48, when it describes the exposed terminal problem, the text and the associated figure 3.3 are inconsistent and confusing.
Chapter 5 discusses a cross layer design for improving the performance of a MANET. A suggestion is made that if the alternative network uses TCP/IP, which has 5 layers, then by modifying some of these layers, efficiencies could be achieved. The discussion does include a frank assessment of possible drawbacks, especially increased complexity of code. If you find the chapter intriguing, there is an entire book that argues for the use of the Internet Protocol, and not for its modification,Interconnecting Smart Objects with IP: The Next Internet by Vassuer and Dunkels. Cross layering can be made to work, but you should first contemplate carefully the chapter's arguments and those of the latter book.
Chapter 12 on mobile data mining on small devices uses Web Services. If you already are familiar with using the Sun Java Wireless Toolkit to program a wireless client (like a mobile phone), then this chapter maybe be very germane, and an easy read.
The problem of malicious nodes in a MANET is very difficult to handle. Chapter 16 uses digital signing of messages between the nodes. So that messages can be authenticated. This has a greater computational load, and hence energy drain, than a shared symmetric key between adjacent nodes. But the latter can be subverted by rogue nodes. Hence the tradeoff, which is not cost free.
If you want an instance of a Finite State Machine, see Chapter 20. It has a nice explanation of VoiceXML, which is an XML vocabulary dedicated to using voice recognition in a mobile phone system. Many commercial applications along these lines now exist, because of significant cost savings from being able to automate user queries.
Chapter 21 tackles the semantic gap between easily computable features of a video stream and what humans might perceive as interesting in that stream. Progress is made on reducing this gap, but an active research effort is still needed.
Chapter 22 might be the portion of the book that comes closest to describing a [possible] mass consumer item. In South Korea, the authors built and tested wearable and [hopefully] fashionable computers. This is the only chapter where aesthetics arise. The test examples hint at future products openly worn by people, yet quite different from today's cellphones. But a key commonality is that wireless connectivity is essential, using popular protocols like WiFi, Bluetooth or Zigbee. If you combine this chapter's findings with a few more years of continued miniaturisation and lowered hardware costs, there is potential for a breakout consumer product. As to what that product might be, perhaps that is your job?
Click Here to see more reviews about: Mobile Intelligence (Wiley Series on Parallel and Distributed Computing)
Get 17% OFF
0 comments:
Post a Comment