- Star topology: All nodes connect to a central hub, typically a switch or router. This configuration simplifies fault detection because a single link failure affects only one node, but the hub itself represents a single point of failure. Star networks are common in local area networks within telecom central offices.
- Mesh topology: Every node connects directly to every other node, providing high redundancy and fault tolerance. While mesh networks require extensive cabling and ports, they are used in backbone networks where reliability is critical, such as between core routers in a service provider network.
- Ring topology: Nodes form a closed loop, with each node passing data to the next. Ring networks offer predictable performance and can survive single link failures if they use dual rings like in SONET/SDH. However, adding or removing nodes disrupts the network.
- Bus topology: All nodes share a single communication line. This simple design is cost-effective for small networks but suffers from collisions and a single point of failure at the bus. It is rarely used in modern telecom due to scalability limits.
- Tree topology: A hierarchical combination of star and bus topologies, often used in cable television and fiber-to-the-home networks. Tree topologies allow easy expansion but require careful management of signal amplification and splitting.
- Hierarchical architecture: Networks are organized into layers: access, distribution, and core. Access layers connect end users, distribution layers aggregate traffic, and core layers provide high-speed backbone. This design improves manageability and scalability in large telecom networks.
- Distributed architecture: Control and processing functions are spread across multiple nodes rather than centralized. This reduces bottlenecks and improves resilience. Examples include distributed base station architectures in 4G/5G where processing is split between central and remote units.
- Software-defined networking (SDN) architecture: Separates the control plane from the data plane, allowing centralized network management. SDN enables dynamic traffic engineering and rapid service deployment, which is increasingly adopted in telecom data centers and wide area networks.
Pattern: Three-tier hierarchical network (Access-Distribution-Core)
Components:
- Access layer: Connects end devices (phones, computers) to the network through switches, base stations, or DSLAMs. Often uses spanning tree or link aggregation for redundancy.
- Distribution layer: Aggregates traffic from multiple access switches and provides policy-based connectivity, such as VLAN routing and QoS marking. Routers at this layer implement routing protocols like OSPF.
- Core layer: High-speed backbone that interconnects distribution layers and provides connectivity to external networks. Core routers use BGP for inter-domain routing and MPLS for traffic engineering.
class StarNetwork:
def __init__(self, hub):
self.hub = hub
self.nodes = []
def add_node(self, node):
self.nodes.append(node)
print('Node {} connected to hub {}'.format(node, self.hub))
def send_data(self, source, destination, data):
if source in self.nodes and destination in self.nodes:
print('Forwarding {} from {} to {} via {}'.format(data, source, destination, self.hub))
else:
print('Node not in network')The code defines a simple star network class where all nodes communicate through a central hub. The add_node() method registers a node, and send_data() simulates data forwarding. This illustrates the star topology concept used in many telecom access networks.
- Star topology uses a central hub and is easy to manage but has a single point of failure.
- Mesh topology offers high redundancy through direct connections between all nodes.
- Ring topology provides deterministic performance and dual-ring protection.
- Tree topology combines star and bus, common in fiber access networks.
- Hierarchical architecture separates access, distribution, and core layers for scalability.
- Distributed architecture spreads control functions to improve resilience.
- SDN decouples control and data planes for centralized network management.