Embedded, Everywhere A Research Agenda for Networked Systems of Embedded Computers (2001) / Chapter Skim
Currently Skimming:
3. Self-Configuration and Adaptive Coordination
3. Self-Configuration and Adaptive Coordination
Pages 76-118
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 76... ...
This chapter focuses on mechanisms needed to achieve automatic reconfiguration. In many EmNets, individual nodes will need to assemble themselves into a networked system, find available resources on the network, and respond to changes in their desired functionality and in the operating environment with little human intervention or guidance.1 A set of basic underlying mechanisms will be required to ensure that EmNets are self-configuring and adaptive.
|
From page 77... ...
For these system components, both the need for efficiency and the constraints on how it is achieved will be more severe than is the case for more traditional distributed computing systems. Efficient system designs will exploit highercapacity and resource-rich components where they exist in the overall system and will exploit the redundancy provided by deploying large numbers of inexpensive components.
|
From page 78... ...
The two processes often occur on different time scales. Adaptive coordination tends to take place more quickly than does self-configuration, with a very short lag time between the moment a change is detected in the operating environment and the time the system adapts its behavior.
|
From page 79... ...
take precedence over other forms of network traffic. In sensor networks that might be used for precision agriculture or environmental monitoring, system composition will vary less because the application is more constrained, while more attention must be paid to adapting the nodes' operational parameters to unpredictable and varying environmental conditions.
|
From page 80... ...
Building on the interface concepts of network configuration, wire protocols, and code mobility, this subsection discusses the issues involved in device and service discovery and how they relate to self-configuration. How entities on an existing network communicate is generally viewed as the interoperability problem.
|
From page 81... ...
environment, for example, the Dynamic Host Configuration Protocol (DHCP) is commonly used to hand out addresses to entities that are connected to the network.
|
From page 82... ...
If more than one object has been registered that implements a given interface, then any of the objects can be returned by such a query. In the lint lookup service, services register by their lava language type; they can be returned to any client asking for something that is at least an instance of the requested class (for example, the returned object might be a subclass of the requested class)
|
From page 83... ...
Rather than returning a directory of service locators, however, the lint multicast request returns a reference that implements the interface to a lint lookup service (including the stub code, or driver, allowing communication with the service) that can be used by the service provider (or client)
|
From page 84... ...
Thus, location can be quite useful in optimizing service discovery as well as in connecting the physical and virtual worlds so that security measures in one can be applied in the other. In other types of EmNets, particularly resource-constrained, wireless networks, network organization needs to correspond more closely with geography in order to be efficient in its use of scarce energy resources (since communication over longer distances consumes significantly more energy)
|
From page 85... ...
Other levels of interoperability include transport protocols (e.g., TCP/IP) that permit a sequence of network packets to be generated and reassembled at the other end, as well as remote procedure calls (RPC)
|
From page 86... ...
Each node may want to indicate to the next how to properly interpret and process each data item. The remainder of this section discusses address configuration, wire protocols, and code mobility as illustrative examples of key interface and interoperability concepts.
|
From page 87... ...
By elevating the level at which the common interface is defined, mobile code enables the protocols used by system nodes to be updated over time or modified for specialization or optimization purposes. Mobile code still requires an initial interface agreement regarding how the code will be transmitted and loaded, but given this foundation and a constant physical layer for communication, it provides a graceful upgrade mechanism for network nodes.
|
From page 88... ...
Because the actual form of the bits on the wire is encapsulated in stub code that comes from the service itself, the wire protocol becomes a private matter between the service and the code it hands out. The client can be, in some sense, far more ignorant; rather than needing code that knows how to translate into a common wire protocol, the client needs only the knowledge of which call method to use.
|
From page 89... ...
The negotiation will include presenting the appropriate access privileges for modifying the code to be run on another node. Advantages of Mobile Code Mobile code has many advantages over wire protocols.
|
From page 90... ...
As background for an analysis of research needs related to EmNets, this section provides examples of how adaptive coordination is handled in more traditional systems. The problems addressed are load balancing, ad hoc routing, and TCP's adaptive congestion control mechanism.
|
From page 91... ...
This approach can be viewed as humanassisted configuration of the system; once the administrator adds the system to the physical cluster, the software is able to automatically reconfigure itself to make use of the extra capacity. Ad Hoc Routing In recent years, other forms of adaptive behavior have been explored in networked systems.
|
From page 92... ...
Adaptive Congestion Control in TCP Another form of adaptive behavior has a completely distributed, local nature TCP's adaptive congestion control mechanism. TCP is the transport protocol run in the Internet over the IP protocol.
|
From page 93... ...
Ad hoc sensor networks provide an excellent example of the issues to be addressed. Many applications require the deployment of sensors over wide areas with no predetermined arrangement.
|
From page 94... ...
Heterogeneity Configuration via Mobile Code Given the expectation of a rapid evolution in hardware, networking protocols, and basic networking algorithms in EmNets, an approach to discovery and configuration based on mobile code seems promising. Such an approach allows these components to evolve separately, rather than requiring that the whole EmNet evolve in lockstep.
|
From page 95... ...
Its properties are abstract and implemented by the underlying components. Indeed, one reason for using mobile code is to allow building the network without having information about the individual components.
|
From page 96... ...
However, the controller for the systems might be a more capable, general-purpose computing element that would interoperate with the rest of the enterprise's inventory and control processes and would benefit from the long-term flexibility of using mobile code technology. Discovery Protocols Current discovery protocols, whether based on wire protocols or mobile code, require that the entity entering the network be able to find, either directly or indirectly, the other entities of interest in the local network neighborhood.
|
From page 97... ...
The issue of low-power discovery is key for EmNets with large numbers of small sensor nodes. At this time, low-power discovery emphasizes the assembly of the physical layer at low power.
|
From page 98... ...
. The implementation of the mobile code can change as new hardware, wire protocols, and software services evolve.
|
From page 99... ...
Although some ideas have been developed about notions of trust in principals, it is not clear that mobile code is a principal, or if such code works on behalf of a principal. Indeed, there are cases in which it makes sense to distinguish between different levels of trust how much the code is trusted to be accurate and nonharmful (which can be thought of as trust in the producer of the code)
|
From page 100... ...
If the automatically configured network has some conceptually central place where members of the network are found, what happens when that place fails? The lint system has a reasonably well-specified failure model, covering both the failure of components that are registered with a lookup service and the failure of the lookup service itself.
|
From page 101... ...
This opens up an entirely new area of research that focuses on techniques for restricting the behavior of EmNets within a parameter space that is comprehensible to both humans and machines. Research Issues for Adaptive Coordination Several factors make it unlikely that adaptive coordination in EmNets will be mediated or even aided by human operators.
|
From page 102... ...
The time scale over which the adaptive coordination may need to take place is too short to be open to human intervention; by the time a human operator decides what to do, the environmental factors will have changed in such a way as to require a completely different adaptation. A third factor is that the operators, users, and individuals interacting with EmNets may be untrained in the specifics of the system and should not be expected to understand the technology to the depth that would be required to address adaptive coordination.
|
From page 103... ...
Especially in sensor networks and other systems based on large numbers of inexpensive nodes, some degree of redundancy can be expected. In sensor networks, for example, multiple nodes may provide coverage of overlapping geographic areas.
|
From page 104... ...
It uses randomization to help coordinate system operations. TCP/IP congestion control scales in the sense that the users of a shared, congested resource use signals (dropped packets)
|
From page 105... ...
Some nodes could enter a sleep mode when redundancy is detected, thereby saving power and contributing to longer system lifetimes. Selfconfiguration itself could take competing paths in which mobile code may be distributed at times when, or to locations where, the combination
|
From page 106... ...
At one end of the spectrum are centralized schemes in which individual components are not self-configuring but the overall system is. At the other end of the spectrum are decentralized schemes in which individual components are themselves self-configuring.
|
From page 107... ...
The viability of a centralized versus a decentralized scheme depends on several factors, including the scale of the system and the rate of anticipated change. Central control across a large network may be impossible to implement in a time-bounded fashion.
|
From page 108... ...
Some systems may benefit from decentralized control schemes, which also require further research and analysis. The minimum number of bits that must be communicated to make a reliable decision is unknown for all but the simplest of problems involving more than one sensor node.
|
From page 109... ...
Preliminary progress has been achieved with simple versions of this problem, but a huge problem space remains to be explored. The interaction between a system element and its neighboring elements is not typically considered in control theory but is essential to
|
From page 110... ...
Centralized control rules can be devised for such a group, but the complexity of the decision-making process, even for a relatively small collection of nodes, will demand some decentralization and probably hierarchy as well. Layered control hierarchies are notoriously difficult to optimize, but perhaps by scaling to large numbers designers can reduce the demand for efficient use of the individual components.
|
From page 111... ...
The problem of cooperation thus appears to offer an excellent opportunity for multidisciplinary research; there are probably lessons to be learned from diverse disciplines, with a potentially high payoff. (An example of an area in which multidisciplinary approaches are used is distributed robotics, described in Box 3.5.)
|
From page 113... ...
At the same time, the system must avoid generating false alarms when a particular event has not occurred. Sensor networks can employ a power-conserving hierarchical detection scheme to meet these objectives.
|
From page 114... ...
This approach requires the exchange of very few bits of information per node. Much more precise position estimates can be achieved with a technique called beam forming, in which individual sensors exchange information about detected events and the time they were detected.
|
From page 115... ...
For example, a dense network of less-intelligent sensors deployed in conjunction with a less-dense array of intelligent nodes could capture information on demand for beam forming. Such collaborative processing can be regarded as a further extension of the signal processing hierarchy to multiple nodes, with the collaboration being extremely expensive in terms of energy use but performed only rarely, such that its marginal energy cost may be acceptable.
|
From page 116... ...
Additionally, if a sensor field is put in a somnolent state in which only selected sensors are powered down, total network power savings will be greater if the multiple sensors coordinate their sleep time (requiring synchronization) as opposed to randomly powering down to provide a reduced alert state overall.
|
From page 117... ...
The state of the art in self-configuration is fairly well developed, with well-understood approaches to address assignment, service discovery, and mobile code. However, significant research progress is needed to achieve automatic self-configuration among large numbers of distributed nodes, while still conforming to well-defined trust and failure models, which are critical to embedded systems applications.
|
From page 118... ...
2000. "Directed diffusion: Scalable and robust communication paradigm for sensor networks." Proceedings of the Sixth Annual International Conference on Mobile Computing and Networks (MobiCOM 2000)
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.