Black and white crayon drawing of a research lab
Research

Multicriteria Decision-Making and applications in the Internet of Things (IoT)

Constanta Zoie Radulescu

Multicriteria Decision-Making (MCDM) or Multiple-criteria decision analysis (MCDA)refers to a collection of methods and tools designed to evaluate, rank, and select alternatives when decisions involve multiple, often conflicting criteria. It is widely used in fields like engineering, healthcare, environmental planning, business, and technology, particularly in complex systems such as the Internet of Things (IoT). MCDA or MCDM is a sub-discipline and full-grown branch of Operations Research.

The expansion of the Internet of Things (IoT) has enabled significant advancements in diverse areas such as healthcare, smart cities, and industrial systems. However, the selection of IoT devices, platforms, and applications involves complex decision-making processes due to the wide range of available alternatives and their varying levels of performance, cost, and functionality. Multicriteria Decision-Making (MCDM) offers systematic approaches to address these challenges by evaluating and prioritizing options based on multiple criteria simultaneously.

The multi-criteria decision-making process is a systematic approach used to evaluate and choose the best alternative among several alternatives based on multiple, often conflicting criteria.

Steps in the Multi-Criteria Decision-Making Process are:

  1. Problem Definition. Defining the decision problem including objective(s) (e.g., selecting the best IoT sensor for a smart city application) and the boundaries and constraints (e.g., budget, technical limitations).
  2. Identification of Alternatives. List all possible alternatives or options that could be chosen.
  3. Selection of Criteria. Identify the criteria that will be used to evaluate the alternatives. Criteria can be quantitative (e.g., cost, performance) or qualitative (e.g., user satisfaction, reliability). The criteria can be of maximum type (benefit) or minimum type (cost).
  4. Assign Weights to Criteria. Tu each criterion can be assigned a weight (coefficient of importance) with a value between 0 and 1. Weights can be calculated using weighting methods.
  5. Evaluation of Alternatives. Score each alternative based on its performance against the chosen criteria. This can be done using quantitative metrics or expert judgment or ratings for qualitative criteria.
  6. Application of MCDM methods. Use an appropriate MCDM method to combine the scores and weights for each alternative to produce a ranking or identify the best option.
  7. Sensitivity Analysis. After obtaining results, perform a sensitivity analysis to test how changes in weights or scores affect the ranking of alternatives. This ensures robustness in decision-making.
  8. Selection and Implementation. Choose the alternative that best meets the decision criteria and implement the solution.

IoT systems involve a multitude of alternatives (devices, sensors, protocols, technologies, applications, platforms) and criteria (cost, reliability, energy efficiency, security, compatibility). IoT selection problems often involve trade-offs between conflicting criteria. Selecting a more cost-effective device could compromise performance or security, and prioritizing energy efficiency may conflict with achieving high performance.

IoT environments are dynamic and subject to changing technology, regulations, and user requirements. Multicriteria methods (MCM) can adapt to evolving criteria or incorporate new criteria. IoT selection decisions can involve risks and uncertainties related to data, technology, or external factors. MCM can incorporate probabilistic models and sensitivity analyses to assess and mitigate uncertainties. IoT selection decisions often affect multiple stakeholders (users, manufacturers, regulators).

Depending on the type of decision problem, MCM can be:

  • classification (descending ordering of alternatives),
  • ordinal sorting / classification (membership of alternatives to predefined decision classes, ordered by preferences),
  • clustering (dividing alternatives into groups according to the measure of similarity / preference),
  • choice (selecting the most preferred subset of alternatives).

Another classification is according to the aggregation procedure into: methods based on ranking relations, utility functions, discrimination functions and without functions.

Also, MCDM methods are classified into: MODM (Multi-Objective Decision Making) and MADM (Multi-Attribute Decision Making). The number of alternatives is infinite in the case of MODM and finite in the case of MADM. These classes are not mutually exclusive; some methods can be used in both categories.

Same examples of multicriteria methods used in selection problems in IoT are: SAW (Simple Additive Weighting), ELECTRE (ÉLimination Et Choix Traduisant la Realité), AHP (Analytic Hierarchy Process), DEA (Data Envelopment Analysis), TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution), PROMETHEE (Preference Ranking Organization Method for Enrich Evaluations), TODIM (Interactive and Multi-criteria Decision Making in Portuguese), COPRAS (Complex Proportional Assessment), ANP (Analytic Network Process), VIKOR (VIseKriterijumska Optimizacija I Kompromisno Resenje), MOORA (Multi-Objective Optimization on the basis of Ratio Analysis), DBA (Distance-Based Approach).

The criteria can have different or equal weights. The weights or importance of the evaluation criteria can be determined using multi-criteria weighting methods, which can include: the Entropy method, DEMATEL (Decision-Making Trial and Evaluation Laboratory), AHP, SMART (Simple Multi-Attribute Rating Technique), CRITIC ( Criteria Importance Through Intercriteria Correlation), SWARA (Step-Wise Weight Assessment Ratio Analysis), WASPAS (Weighted Aggregated Sum Product Assessment), BWM (Best-Worst Method).

The field of multi-criteria decision-making has developed a lot in recent years. New methods, adaptations and extensions of existing methods, hybrid methods (combinations of two or more methods) have appeared. Multi-criteria methods have continuously evolved to cope with the diversity and complexity of current IoT decision-making problems.

Our applications of MCDM in IoT problems

An analysis of the use of multi-criteria methods for solving selection problems in IoT systems has highlighted that the most widely used methods are AHP, TOPSIS and SAW, as well as combinations of these methods with other methods. The AHP method is used to calculate the weights of the criteria and is often found in combination with other multi-criteria methods. Both classical and newer methods are used, and many approaches are hybrid or combinations of methods. The selection problems addressed are: selection of devices, sensors, applications, platforms, service providers, communication networks, IoT services for different domains: IoT in industry, supply chain networks, health, cyber-physical systems. The criteria taken into account differ depending on the type of selection problem.

The growth of Internet of Things (IoT) systems is driven by their potential to improve efficiency, enhance decision-making, and create new business opportunities across various domains. A Hybrid Group Multi-Criteria Approach for solving selection problems in IoT-type systems is proposed. The approach contains the Best Worst Method (BWM) weighting method, multi-criteria Simple Additive Weighting (SAW), Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS), VIseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR), and Complex Proportional Assessment Method (COPRAS), and a method that combines the solutions obtained using the four considered multi-criteria methods to obtain a single solution.

The SAW, TOPSIS, VIKOR, and COPRAS methods were analysed in relation to their advantages, disadvantages, inputs, outputs, measurement scale, type of normalization, aggregation method, parameters, complexity of implementation, and interactivity. An application of the Hybrid Group Multi-Criteria Approach for IoT platform selection and a comparison between the SAW, TOPSIS, VIKOR, and COPRAS solutions and the solution of the proposed approach is realized. A Spearman correlation analysis is presented.

The diversity of Internet of Things (IoT) systems can be viewed from two points of view, one that consists of providing the necessary applications and another that can lead to a large number of security threats and attacks. The analysis of the influence between different IoT security requirements (IoT-SR) as well as the determination of the importance of each security requirement play a vital role in the issue of the effective evaluation of an IoT system. The diverse nature, importance, evaluation and influence of multiple IoT-SR are the main issues that make this problem a multi-criteria problem. Considering the qualitative nature of IoT-SR and the advantages of fuzzy sets in efficiently dealing with uncertainty, fuzzy subjective weighting methods can be used to address the problem. A Multi-Criteria Weighting Approach (MCWA) based on the Fuzzy DEMATEL method in combination with two weighting methods, aiming to analyze cause-and-effect relationships for a set of criteria and to calculate the criteria weights is proposed. An application of MCWA for IoT-SR is realized. The obtained results can open new ways for system security specialists to focus on security requirements determined to be of increased importance. Prioritization of the IoT security requirements represents an important element in the security evaluation of these systems.

A modification of the TOPSIS method by using another way of calculating the relative closeness coefficient (RCC) and taking into account a group of decision makers is proposed. The modified version of the TOPSIS method, call the Linear Trade-off Group TOPSIS method (LTG-TOPSIS), replaces the RCC from the classical TOPSIS method with one that depends on a parameter that takes values in the unit interval. By using the parameter, the RCC is calculated as a linear combination between distances of an alternative to the ideal and anti-ideal solutions. This approach facilitates a management of the compromise between the two distances. An implementation of the LTG-TOPSIS method is analysed for a IoT devices ranking. By varying the parameter of the proposed method, a set of IoT devices rankings is obtained and the change in the ranking is studied. A comparison of the rankings obtained with LTG-TOPSIS and with the classical TOPSIS method is performed.

Sensors for health are a dynamic technology and sensor-based medical devices (SMD) are becoming an important part of health monitoring systems in healthcare centers and ambulatory care. The rapid growth in the number, diversity and costs of medical devices and Internet of Things (IoT) healthcare platforms imposes a challenge for healthcare managers: making a rational choice of SMD vendor from a set of potential SMD vendors. A hybrid approach that combines a performance evaluation model and a multi-objective model for the SMD vendor selection problem is proposed. For determining the criteria weights in the performance evaluation model, an original version of the best worst method (BWM) is applied, which we call the flexible best worst method (FBWM). The multi-objective model has two objective functions; one is to maximize the SMD performance and the other is to minimize the SMD cost. A case study for the application of the hybrid approach for SMD procurement in a healthcare center is analysed. The hybrid approach can support healthcare decision makers in their SMD procurement decisions.

Each multi-criteria method has advantages and disadvantages when used in selection problems. The choice of method depends on the problem, the available data, and the preferences of the decision makers. Evaluation and selection problems in IoT differ from general selection problems due to the unique characteristics and challenges associated with IoT systems. These problems involve a unique set of considerations related to complexity, heterogeneity, scalability, security, data quality, and interoperability. IoT systems are complex, involving multiple and different devices and sensors that use different protocols and standards; complexity and heterogeneity make it difficult to evaluate alternatives and select the best solution. The scalability of IoT (by continuously adding new elements) increases the difficulty of selecting a valid solution for future developments. Security must be carefully considered, by protecting sensitive data and systems from unauthorized access, theft, or damage. IoT devices may be vulnerable to cyberattacks due to their increased exposure to the Internet and limited computing resources. IoT systems generate large amounts of data, which are ‘noisy’, incomplete or inconsistent. Data quality and security must be ensured in the selection of the best solution and in making relevant decisions. The selection process should be scalable, adaptable to different dimensions of IoT systems. These systems are subject to regulations and standards related to data confidentiality, security and environmental factors, and selection decisions must align with legal and compliance requirements. The complexity, multi-dimensionality and importance of selection decisions in these systems make multi-criteria methods a valuable tool for solving selection problems because they allow decision-makers to control the complexity of IoT environments, make informed choices and optimize resource allocation to achieve their objectives effectively.

Multi-criteria (multi-attribute) methods are useful in IoT because:

  • they help to manage the complexities of IoT problems by evaluating multiple options based on a set of relevant criteria;
  • they allow for a holistic evaluation of IoT solutions, simultaneously considering factors such as technological performance, costs, security, interoperability, and sustainability;
  • they can be customized according to the specifics of the IoT problem, depending on the priorities and objectives of the organization implementing the solution;
  • they provide a rigorous framework for evaluating and selecting optimal IoT solutions, which is essential in the context of complex decisions in this industry.

Read our papers:

  1. RADULESCU, C. Z., & NEACȘU, D. M. (2023). Selecția în sisteme de tip IoT bazatӑ pe metode multi-criteriale. Romanian Journal of Information Technology and Automatic Control, 33(3), 113-128. https://doi.org/10.33436/v33i3y202309 https://rria.ici.ro/en/vol-33-no-3-2023/selection-in-iot-type-systems-based-on-multi-criteria-methods/
  2. Radulescu, C. Z., & Radulescu, M. (2024). A Hybrid Group Multi-Criteria Approach Based on SAW, TOPSIS, VIKOR, and COPRAS Methods for Complex IoT Selection Problems. Electronics, 13(4), 789. https://www.mdpi.com/2079-9292/13/4/789
  3. Radulescu, C. Z., Boncea, R., & Vevera, A. V. (2023). A Multi-criteria Weighting Approach with Application to Internet of Things. Studies in Informatics and Control, 32, 4. https://sic.ici.ro/vol-32-no-4-2023/a-multi-criteria-weighting-approach-with-application-to-internet-of-things/
  4. Radulescu, C. Z., Radulescu, M., & Boncea, R. (2023, May). An Application of the Flexible Best–Worst Method to Weighting Internet of Things Security Requirements. In International Conference on Informatics in Economy (pp. 207-218). Singapore: Springer Nature Singapore. https://link.springer.com/chapter/10.1007/978-981-99-6529-8_18
  5. Radulescu, C. Z., Radulescu, M., & Boncea, R. (2024). A Linear Trade-off Group TOPSIS method with application for Internet of Things devices ranking. Procedia Computer Science, 242, 528-535. https://www.sciencedirect.com/science/article/pii/S1877050924018180
  6. Radulescu, C. Z., & Radulescu, M. (2023). A hybrid multi-criteria approach to the vendor selection problem for sensor-based medical devices. Sensors, 23(2), 764. https://www.mdpi.com/1424-8220/23/2/764
Go back