The routing protocol for low-power and lossy networks (RPLs) is the routing layer protocol for IoT. The RPL is a proactive routing protocol for multi-hop communication. It is designed based on distance vectors, and it is operated on IEEE 802.15.4 MAC protocol. The RPL constructs the topology structure similar to a tree named Destination Oriented Directed Acyclic Graph (DODAG) and eases the packet routing. It takes one or more routing metrics to estimate the rank for DODAG construction. From rank estimation to topology construction, RPL uses various control messages.
Three control messages are defined as follows: • DODAG Information Solicitation (DIS): Used to respond to connected neighboring nodes about DODAG joining. • DODAG Information Object (DIO): DIO message periodically refreshes the information of the nodes about the version of DODAG, RPLInstanceID, DODAG version number, and DODAG ID. • Destination Advertisement Object (DAO): DAO helps build routes from root nodes to sensors.
• The RPL accounts for the Objective Functions (OFs) with link quality measurements and rank estimation to construct the DODAG to root devices separately and trickle algorithm. Recent researchers mainly focus on those techniques for enhancing the efficiency and security of routing layer protocol.
Project Ideas in RPL Routing
Cooja Projects based on Objective Functions in RPL Routing Protocol • The RPL accounts for the Objective Functions (OFs) with improved link quality measurements and rank estimation to construct the DODAG to root devices separately and trickle algorithm. Recent researchers mainly focus on those techniques for enhancing the reliability of routing layer protocol.
Critical Analysis of Objective Functions: • To understand the protocol behavior for different individual and combined metrics, it is prominent to evaluate the RPL over different scale and mobile environments. • A structured algorithm is needed for Objective Function (OFs) and Minimum Rank Hysteresis Objective Function (MRHOF). • With the knowledge of routing path cost evaluation using different metrics, a generalized algorithm is developed to enhance RPL performance.
RPL objective functions using variant routing metrics: • ETX and Hop Count alone are insufficient to cover all the IoT needs. • According to the application requirement, the routing metric selection is a principal idea to improve the RPL performance. • A combination of three different metrics ETX, Content, and Energy related metrics, are used to enhance the design of the objective function of RPL for IoT applications. It explores the enhanced triggering technique in the designs of improved RPL.
Additive Combination of Node and Link Metrics: • Default OFs in RPL decide the routing paths according to a single routing metric, and it may be an unoptimized path and many parent changes under mobile and large-scale networks. • An optimal way to design optimal PRL is to account for weighted combined metrics objective function (WCM-OF) and non-weighted combined metrics objective function (NWCM-OF). They group routing metrics such as link quality and energy with equal weights.
Energy Retaining Objective Function in RPL Routing Protocol: • Different OFs can be developed for RPL. Most of the default schemes fail to consider the remaining energy, leading to imbalanced DODAG construction. • Different metrics compositions should be evaluated to attain reliable communication and maximize the network lifetime.
Intelligent Learning Automata Based Objective Function in RPL Routing Protocol: • A dynamic and lossy IoT environment challenges the RPL routing in the network layer. • To improve the RPL performance, learning automata is applied to tune ETX as per the network conditions.
Clustering Model for RPL Routing Protocol Enhancement
Cluster Parent Based RPL Routing Protocol: • An opportunistic coordination forwarding scheme is developed using an optimal cluster-parent-based RPL scheme. • The usage of a priority-based scheduling mechanism in RPL reduces the end-to-end cost of DODAG construction without increasing the energy consumption.
Load Balanced Clustering Method for RPL Routing Protocol: • The cluster ranking-based method named C-balance improves the DODAG efficiency and RPL performance. • Multiple routing metrics are involved, such as Expected Transmission Count, hop count, residual energy, and the number of children.
Energy Efficient RPL Routing Protocol
Context-Aware RPL Routing: • Hop count and ETX are commonly considered routing metrics for DODAG construction. However, they are insufficient to design an optimal route. • Context-aware RPL is essential to account for energy and traffic-related metrics in route construction and efficient RPL routing.
Load Balancing RPL Routing: • In RPL DODAG construction, communication quality-related metrics should be considered to optimize the network's lifetime. • Some of them are ETX, power consumption (PC), and the energy balancing factor. It is essential to consider the traffic-aware metric that utilizes parent count metrics (ETXPC) for ETX rather than individual ETX.
Multipath RPL Routing: • To enable multipath routing OFs, more than one parent node should be selected to reduce the burden on the same parent node and equalizes the energy consumption. • Packet replication and elimination are introduced. It takes multiple copies for each message, and they are transmitted in parallel along different routes to improve communication reliability.
Compression Aware Aggregation based RPL Routing: • RPL routing is designed along with the compression and aggregation concepts to improve the efficiency of RPL under an IoT environment. • Each region has one aggregator and is selected based on ETX and energy factors. A compressor helps compress the aggregated messages and improves the RPL communication efficiency.
Scheduling Aware Forwarding Mechanism in RPL Routing: • To improve the network lifetime while improving the load balancing, scheduling-aware RPL is designed. • Scheduling matrix information is designed to know the occupied cells and take an efficient routing decision under a lossy and dynamic environment.
Scalable and Mobility Aware RPL Routing
Scalable RPL Routing: • Scaling the RPL routing to dense and large area networks with limited battery and memory resources improves the RPL processes. • By maintaining a stable and reliable DODAG structure and appending a novel policy for managing the neighbor table, the RPL performance can be improved.
Impact of Mobility Models on RPL Routing: • Several mobility models have been designed to mimic the mobility of different devices under different IoT applications. • The impact of those models on RPL routing performance metrics is evaluated to analyze the RPL for further enhancement.
Reliable and Mobility Aware RPL Routing: • A dynamic and lossy environment makes frequent changes in DODAG structure and reduces the performance of RPL under an IoT environment. • It necessitates the design of a stochastic parent replacement policy to solve such issues. The mobility metric, Time to reside, involves improving the RPL performance under the mobility environment.
Advanced Mobility Support Routing Protocol: • A new proactive mobility support protocol based on the RPL standard is designed using an Extended Kalman Filter. • It can offer RPL communication with seamless connectivity while reducing the number of switching between parent nodes in the DODAG structure.
Neighbor Variability based RPL Routing for Mobile IoT: • Neighbor variability detection by comparing the current neighbor list with the previous one, the impact of mobility is identified. • It can prevent the DODAG disconnection and improve the RPL performance in a dynamic and lossy environment.
Optimization Scheme Based RPL Routing Enhancement
Game-Theoretic Optimization for RPL Routing: • Game-theoretic model is formulated among IoT nodes, and it models the scenario where nodes compete for network resources. • Different routing control parameters are considered in the game model to take an efficient RPL decision and improve the performance metrics.
Novel RPL Routing Based on Chaotic Genetic Algorithm: • Instead of considering a single routing metric, recent research mainly focuses on multiple metrics and their composition technique to improve and maintain the RPL performance. • A chaotic genetic algorithm helps identify the better weighting distribution of every routing metric in the composition metric and select an efficient parent node to form the DODAG construction.
Fuzzy Logic Based RPL Routing: • A fuzzy logic-based routing decision plays a crucial role in building a load-balanced network using an evenly distributed load and energy consumption among the nodes. • It is essential to incorporate multiple load-related routing metrics in fuzzy logic based RPL routing protocol
Reinforcement Learning based RPL Routing: • Cooperation among different layers is important to improve the IoT wireless communication efficiency. • The routing layer explores the ranking mechanism for packet routing. Q-learning is an efficient way to identify the collision-free DODAG structure and efficient routing protocol design.
Cross-Layer Based RPL Routing
Elaborated Cross-Layer OFs in RPL Routing: • An energy-efficient cross-layer OFs is designed to improve the RPL routing under IoT. • It explores a novel routing metric, named strobe per packet ratio, which defines the number of transmitted strobes per packet due to Radio duty cycling. • It decides the MAC layer protocol and improves the performance using a cross-layer approach.
Opportunistic RPL Routing: • Improving the quality of communication over unreliable links and improving packet delivery without a deadline while minimizing overhead and energy consumption using RPL is essential. • The QoS communication is achieved by extending RPL to opportunistic routing instead of single-parent based unreliable communication.
Congestion Aware RPL Routing Protocol: • Most of the previous works focus on energy balancing and communication delay, regardless of the network traffic and congestion. However, network congestion is an important problem that needs to be solved in the IoT environment. • Recently, RPL researchers have focused on improving the congestion control schemes by considering network queue, message transmission duration, and packet loss.
Trickle Improvement Schemes for RPL Routing Protocol
Efficient Parent Selection and Dynamic Trickle Algorithm: • ACO-based multi-factor optimization is designed to select an optimal DODAG structure, and the coverage-based dynamic trickle algorithm reduces the routing overhead without affecting its routing performance. • It considers ETX, hop count, and children count for rank estimation. The concentric corona-based decision on control message broadcasting using the Trickle algorithm mitigates the control overhead effectively.
Novel Flexible Trickle Algorithm: • The flexible trickle algorithm takes an efficient decision on control message broadcasting based on the transmission time and the data delivery intervals. • The comparison of FL-trickle with a benchmarked model demonstrates that FL-trickle outperforms the basic Trickle algorithm in terms of delay, overhead, energy consumption, throughput, and network lifetime.
Ideas of Novel Trickle Timer Algorithms: • A main limitation of the basic Trickle algorithm is it suffers from the short listen problem. • Different works are developed to dynamically improve various trickle parameters to limit control overhead.
Dynamic Redundancy based RPL Protocol: • Issues of basic Trickle algorithm are evaluated under even and uneven node distributed environments.
• The dynamic trickle algorithm adopts the redundancy parameter in the Trickle algorithm and tries to reduce the overhead with even distribution of energy resources.
Cooja Projects on RPL Security
Attacks in RPL Routing: • An attacker can launch several attacks against IoT devices to interrupt the network. Routing attacks are most common in low-power wireless networks. • The RPL attacks can be classified according to the network topology, traffic, and resources. • Network Topology related attacks are Black-hole, Worm-hole, sink-hole, DAO inconsistency, Worst parent, Replay, and Routing Table falsification. • Traffic-related RPL attacks are Traffic analysis, identity attack, decreased rank attack and sniffing. • Network resource-based RPL attacks are Flooding, routing table overload, DAG inconsistency, and Version number attacks.
Trust-based security Scheme
Trust Model for Selective Forwarding Attack Detection: • A new centralized trust-based defense mechanism is designed to combat selective forwarding attacks in RPL. • It considers the DODAG structure, i.e., RPL network topology, for detecting the malicious nodes from the normal ones.
Lightweight Trust based Model: • The lightweight trust-based model incorporates direct and indirect trust estimation from direct traffic observation and recommendation mode, respectively. • It appends the received signal strength along with the trust value in rank estimation and builds a lightweight and secure DODAG structure for RPL communication.
Control Layer Based Trust Model: • The forwarding behavior of nodes is observed, and the trust value is estimated for every node. • Instead of estimating the trust values on the controller, selected nodes are involved in the trust estimation. • It avoids high energy depletion and poor network lifetime without compromising network security.
Trust-based Optimized RPL: • Most of the previous works perform node localization and routing separately. • Trust-based optimized RPL locates the trusted nodes in routing paths using dragonfly optimization. • Trust and optimization scheme improves the network lifetime and communication security.
SecTrust-RPL Model: • It embeds the secure trust model in RPL functions to identify the rank and Sybil attacks. • Direct and recommended trust estimations are involved. Both periodic and reactive trust update models are incorporated as per the network scenario.
Intrusion Detection Systems
Anomaly Based IDS Model: • Anomaly-based IDS explores the normal behavior of a node as a baseline. Using such normal behavior profile, it can detect anomalies when there are deviations from the baseline. • To recognize the normal routing activities, any learning model can be used. • It can detect new attacks, but there may be a possibility for high false positive and false negative rates, especially when they show small deviations from the baseline.
Signature Based IDS Model: • Signature-based IDS explore predefined attack patterns, named signatures. • This approach can detect defined routing attacks and secure and trusted parent nodes among neighboring nodes. • It helps build a secure DODAG structure and secure communication against RPL routing attacks.
Strainer-based intrusion detection: • The blackhole attacks in RPL networks advertise a greater routing metric to neighboring nodes, and it is selected as a preferred parent. • Strainer-based IDS uses the point above to identify the suspicious nodes. The IDS is responsible for analyzing those nodes' behavior and isolating the confirmed malicious nodes from the network.
Sink based IDS: • A sink-based IDS model is used to detect decreased rank attacks in non-storing RPL networks. • The root node is responsible for detecting and isolating the network's malicious nodes. • It stores and compares the previous and current ranks of parent nodes and detects a threshold value for parent switching. If a node crosses the threshold, the sink or root node marks it as malicious.
Threshold Based IDS: • The threshold-based IDS explores the Sequential Probability Ratio Test (SPRT) along with an adaptive threshold. • There are two modules: decision-making at the root node and selected nodes monitoring the network traffic. • The monitoring nodes send information to the root via randomly selected paths. The root node is responsible for analyzing the data to estimate the probability of a node being malicious.
Optimization and Learning Model Based RPL Security
Machine Learning-Based RPL Security: • Machine Learning-based secure RPL routing (MLRP) protocol helps analyze the RPL traffic information and identify the malicious packets under different classes accurately. • A complex dataset with malicious and normal network behavior is generated for RPL protocol using the Cooja simulator.
Game Model for RPL Security: • IoT routing and its features can be modeled as the game scenario, where nodes compete for network resources to send their data packets to the sink node. • A game model helps analyze the interactions with attackers and take the best optimal prevention action with the least cost. • It can offer an energy-efficient RPL security scheme.
Fuzzy Logic For RPL Security Provisioning: • Local Repair attack on RPL DODAG impacts the security and efficiency of RPL routing under IoT environments. • The fuzzy logic method permits the security scheme to convert multiple input variables into one output and identifies the attacker nodes using the steps of fuzzification, fuzzy inference, aggregation, and defuzzification.