The importance of anchor
Linear anchors the network identification may have many uses, for example in the community with a linear broadband network topology in which the anchor can be used as a gateway to the Internet, to optimize the overall transmission time of the network community.
When used for military or emergency response scenarios, one of the anchors by the central node to add as LL, it is possible to create a network topology having a small APL value.
Identifying anchor also facilitates vehicle communication.
For a given graph, minimizing FIG APL is equivalent to minimizing the total path length.
A fixed percentage of the anchor position is always 0.2N or 0.8N.
Add a link based on deterministic heuristics of
two kinds of certainty link to add the policy, ie the maximum difference between CC (MaxCCD), and the order of certainty LL added.
Close to the center of the difference between the two nodes CCD defined as the difference between the values of the two CC nodes.
MaxCCD strategy to add LL between nodes having the largest of the CCD.
APL represents an average path length between a pair of nodes throughout the network.
AEL characterize the average length of each link of the network.
BC value of the node indicates its importance in the network.
CC value of the node characterizes the closeness of the node and other nodes.
Average network delay: (Average Network Delay, ANeD) measures a set of data propagates from the source node to the destination node the average time required. ANeD transmission delay equal to the propagation delay and sum.
L is deterministic order to add another addition method is determined based heuristic LL, transforming it by regular linear network strip LL k small-world network configuration.
The average flow capacity based Small World features enhanced heuristic algorithm ACES
Brace paradox
Small World routing network
routing may be defined as a network specific information is forwarded from the source node to the destination node process.
Distributed routing algorithm
adaptive distributed routing algorithm
proactive routing algorithm
Small world network capacity of
the network capacity is defined as part of the network may be transmitted from another portion of the amount of information per unit time.
Increase network capacity to improve the overall performance of the underlying network is one of the key challenges.
It can be converted to a regular network of small-world network transformation in two ways: to reconnect existing link NL; add a new link LL
Chapter V scale-free networks
Widely present in nature in scale-free networks follow a power law degree distribution.
The method of creating a variety of scale-free network: connection preference; by adaptation of the model; by changing the inner fitness; and by the popularity of local similarity optimization; use index 1; greedy global optimization.
Simple greedy local optimization may lead to a high degree of some node goes popularity and similarity.
Characterized in scale-free networks: making complex networks become the main features of scale-free networks: presence of central node; small world attributes; ultra-small diameter network; degree distribution power; adjacency matrix based on the Laplace and spectral matrix distributed.
1. The existence of the central node of
scale-free networks resistance random attack is very strong, however, in the face of a targeted attack and is very fragile.
2. small-world properties
of the ACC property can be identified by the average path length a low APL and a low to moderate average clustering coefficient.
APL can be defined from the average number of hops in the network end nodes among all pairs. Progressive scale-free networks APL value O (loglogN), N is the size of the network.
ACC characterizes the network each node is the average number of neighbors each other neighbors.
3. Network diameter of
the smaller diameter of the network is scale-free networks Another key feature. The diameter of the network can be defined as the distance the longest possible network end. Scale-free networks, also known as ultra-small network.
4. The distribution
of the network distribution reflects the overall condition of the network communication. Distribution statistics of the number of nodes in the network have the same degree. Also referred to as a power law long-tailed distribution, wherein some nodes have a very high degree, while the rest of the nodes is relatively low.
5. FIG spectral distribution of
the spectral distribution of scale-free networks has a triangular shape.
Formation of scale-free network
during network evolution, the following mechanisms may form a scale-free networks: growth and preferential attachment; Preference connection based adaptation degree; the degree of variation inherent in the adaptation; similarity and popularity optimization; network evolution of global optimization based on greedy; index having a scale-free networks.
1. Create a preferential attachment scale-free networks: the new node network connected to those more inclined to a higher degree nodes.
2. fitness model created by scale-free networks: a node or several nodes become very popular in a short time;
3. Create scale-free networks through internal variable Adaption: new node based on the node connected to certain characteristics of existing nodes in the network, and that the benefit of both nodes to each other;
4. optimization of scale-free networks created by: similarity measures the proximity of nodes in the network with the newly added node, If the new opportunity that is closer to the network node in a node, the new node is connected to the existing nodes even higher.
The scale-free networks created by index 1: r = to generate an unscaled model by application of the concept of the random mapping process.
6. Create scale-free networks by greedy global decision-making: After you add a group of greedy LL through a series of global decision-making, small-world networks can be transformed into scale-free networks, LL obey remote link affinity LRA here newly added constraints. Betweenness centrality described the case where the network node is the shortest path between the different nodes, the higher the number of central mediator node, the remaining nodes to establish a new connection greater affinity, in turn, can evolve into a central node in the network.
7.BA network evolution model by preferential attachment and growth of new network links in combination.
8, create a variable based on internal fitness of scale-free network
connection is established between these nodes with each other on a reciprocal basis.
Popularity is similar appeal, it may be considered as a potential preference for power connection; on the other hand, reflects the similarity network groups with similar interests. Homogeneity (homophily)
Scale-free networks are also known as ultra-small-world networks.