13.07.2015 Views

Page 2 Lecture Notes in Computer Science 2865 Edited by G. Goos ...

Page 2 Lecture Notes in Computer Science 2865 Edited by G. Goos ...

Page 2 Lecture Notes in Computer Science 2865 Edited by G. Goos ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

230 F.J. Mol<strong>in</strong>a, J. Barbancho, and J. LuqueTelemetry services:– TLM Poll<strong>in</strong>g. This service <strong>in</strong>itiates a periodic poll<strong>in</strong>g to a subset of nodes,located at distribution po<strong>in</strong>ts. It creates and updates a table called ImageTable with data from sensors which <strong>in</strong>clude a timestamp. This table may beaccessed <strong>by</strong> custom primitives (TLM Read ).– TLM Read. Return data from image table with <strong>in</strong>tegrity <strong>in</strong>formation (timestamp).Remote control services:– RC Send. Root sends a message with orders to control remote device (valves,breakers...).Message must be sequenced to avoid duplication and encryptedfor secure operation. It may be necessary to notify a receipt message, andan order completion message.– AMR Collect service may be the most complex because time m<strong>in</strong>imizationand network overhead reduction are quite difficult to optimize simultaneously.There are many studies for poll<strong>in</strong>g optimization [17,18], but most ofthem use a well def<strong>in</strong>ed topology (rectangular, hexagonal...). The proposednetwork is a random network, we only suppose that nodes are uniformlydeployed either as short-range nodes or long-range nodes. Execution timeand network overheads can be evaluated approximately from topology parameters(see previous section). These values may be used to compare howdifferent poll<strong>in</strong>g schemes may optimize network performance.From the medium number of transmissions, we can compute execution time<strong>in</strong> an ideal context (equation 4) or <strong>in</strong> a more real one (equation 5). We canconsider the collect message like a token mov<strong>in</strong>g along the branches from rootto nodes, and the answer as another token return<strong>in</strong>g back from nodes. The firstapproach is a simple collect<strong>in</strong>g schedule consist<strong>in</strong>g of poll<strong>in</strong>g each node from theroot and, <strong>in</strong>dividually, wait<strong>in</strong>g for the answers. Nodes do not have an applicationlayer <strong>in</strong> that case, and there is only a s<strong>in</strong>gle token mov<strong>in</strong>g <strong>in</strong> the tree. Moreover,let us consider an ideal scenario with the follow<strong>in</strong>g conditions:1. There is no <strong>in</strong>terference or collisions.2. Wait<strong>in</strong>g time to medium access is zero.3. Protocol stack comput<strong>in</strong>g time is negligible.4. Only one frame is necessary to transmit all data from a node.5. All tokens have exactly the same size.An approach value of execution time, for an overall collect order, may bederived from a medium number of transmissions (equations 2,3,4), as:CT 1 = ( PTS · N · NT +ATS· N · NT ) ·CharSizeBaudRate(7)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!