Sensing as a Service

• Sensing as a Service is a paradigm that transforms traditional sensor deployments into brokered, on-demand, pay-per-use resources accessible to multiple consumers.
• It leverages layered architectures with Sensor Publishers and Extended Service Providers to enable dynamic discovery, flexible pricing, and multi-tenant access.
• The model integrates economic mechanisms, robust security protocols, and regulatory safeguards to ensure scalable, efficient, and compliant sensor data marketplaces.

Sensing as a Service (SenaaS) is a layered, market-driven paradigm that transforms physical sensor deployments into dynamic, brokered, pay-per-use resources to be discovered and consumed on demand across multiple tenants. Originating at the intersection of Internet of Things (IoT), cloud computing, and smart city initiatives, SenaaS reconceptualizes sensors as rentable entities, accessible not just to their owners but also to diverse third-party consumers, through discoverable registries, automated negotiation, and scalable data marketplaces (Perera et al., 2013). The model leverages contemporary IoT networks, low-power protocols, and cloud-based middleware, incorporating advanced architectural features, economic models, and social–regulatory safeguards to address the requirements of a massive, heterogeneous, and economically viable sensor ecosystem.

1. Conceptual Foundations and Model Differentiation

SenaaS operates by exposing physical sensors—ranging from embedded devices in appliances, vehicles, and city infrastructure to dedicated sensor networks—through intermediaries known as Sensor Publishers (SPs). Sensor owners, spanning personal, commercial, and public domains, register sensors with SPs, who then publish metadata-rich sensor profiles to discoverable catalogs and negotiate access with Sensor Data Consumers (e.g., governments, research organizations, or businesses). A unique differentiator is the support for dynamic brokered access: rather than static, one-to-one integrations, sensor resources are matched to arbitrary consumer requests in real time, with flexible pricing and access policies (Perera et al., 2013).

Key model features include:

• Multi-tenancy: A single sensor can be accessed by multiple independent consumers under negotiated terms.
• Market-based access: Sensors are subject to pricing and SLA negotiation, frequently via marketplace or auction mechanisms.
• Owner-driven control: Owners approve each access transaction, with explicit policy setting for privacy, quality, and return.
• Support for extended service layers: Value-added intermediaries, called Extended Service Providers (ESPs), aggregate, curate, or analyze raw data streams, further decoupling data ownership from value creation.

This model stands in contrast to traditional IoT deployments, which usually entail hardwired, application-specific sensor installations and cloud models serving as either vendor-specific silos or large-scale infrastructure services with limited openness to external consumers.

2. SenaaS Architectural Layers and Workflow

The canonical SenaaS architecture comprises four conceptual layers (Perera et al., 2013):

1. Sensors & Sensor Owners: The foundational layer, encompassing all physical devices with sensing capability. Owners segment into personal, private, public, or dedicated commercial categories. Sensor registration includes computation of quality attributes, pricing bounds, privacy guards, and participation policies.
2. Sensor Publisher (SP): Agents that manage the discovery, aggregation, and publication of sensor metadata into a registry. They match consumer requests to available sensors and forward access offers to the respective owners, who may approve or deny access on a per-request basis. Post-approval, SPs manage streaming, SLA enforcement, and billing.
3. Extended Service Provider (ESP): Middleware or brokers that coordinate across multiple SPs to fulfill complex or multidimensional consumer requirements (e.g., multi-modal environmental indices), offer advanced analytics, provide semantic discovery, and ensure SLA management. ESPs receive commissions for orchestration.
4. Sensor Data Consumers: Entities that articulate high-level query requirements, including spatiotemporal bounds, data quality, and budget. Consumers may interact directly with SPs or use ESPs for comprehensive end-to-end service.

Data flows from sensor to consumer through the following stages: discovery and registration, metadata cataloging, request–offer matching, owner approval, streaming under SLA negotiation, and financial settlement.

3. Technological Enablers and Infrastructure

SenaaS builds atop established IoT and cloud technology stacks (Perera et al., 2013). Key enablers include:

• Device protocols: Lightweight standards such as CoAP, MQTT, 6LoWPAN, ZigBee, and BLE optimize communication for energy and bandwidth-constrained sensors.
• Connectivity: A spectrum of wide-area technologies (Wi-Fi, 3G/4G/5G, LPWAN) supports deployment across varied urban and rural contexts.
• Cloud APIs and middleware: RESTful endpoints and real-time streaming (e.g., via WebSockets), cloud platforms like Xively or the Global Sensor Networks (GSN), and semantic provisioning middleware (e.g., OpenIoT) manage device integration, metadata, and access control.
• Orchestration frameworks: Docker and Kubernetes facilitate modular scaling and lifecycle management of SP/ESP brokers.
• Performance and aggregation strategies: Data streams are subject to local preprocessing, sliding-window aggregation, and application of cost-aware sampling heuristics to optimize bandwidth, energy, and SLA compliance. Quality-of-Service (QoS) is formalized via metrics for latency, throughput, and availability, commonly with SLA requirements at or above 99.9% uptime.

4. Economic and Business Model

SenaaS establishes a multi-sided digital marketplace with distinct roles and revenue flows (Perera et al., 2013):

• Sensor Owners monetize underused sensors, receiving per-use or subscription-based compensation.
• Sensor Publishers and ESPs earn commission-based revenue for matching, brokering, or value-adding.
• Consumers are charged via flexible models: pay-per-use, subscription, or transaction-based pricing, with opportunities for customized SLA terms.

Transaction workflows incorporate negotiation: consumers propose budgets or bid prices, while owners counter with their acceptable terms, including monetary and non-monetary returns. Revenue-sharing among intermediaries typically follows commission rates, as captured by formulas such as

$$\text{Revenue}_{SP} = c_{SP} \cdot \text{Price}_{transaction}, \quad \text{Revenue}_{ESP} = c_{ESP} \cdot \text{Price}_{transaction}$$

where commission rates $c_{SP}, c_{ESP} \in (0,1)$ are negotiated per arrangement.

ESPs may provide differentiated services—such as guaranteed data curation quality, advanced anomaly detection, or unique semantic fusion—to justify additional fee layers.

5. Social, Legal, and Regulatory Framework

SenaaS is underpinned by robust privacy, security, and governance regimes (Perera et al., 2013):

• Data ownership and explicit consent: Owners retain control, approving each sensor’s registration and publication, and defining fine-grained access policies.
• Anonymization and privacy preservation: Data streams are processed to strip or generalize personally identifying features, leveraging hardware-level anonymizers and middleware protocols (e.g., for k-anonymity).
• End-to-end security: Encryption is implemented using DTLS (for CoAP) or TLS (for MQTT/HTTP), with device attestation and credential management.
• Governance and compliance: Clearly defined liability boundaries, data breach notification protocols, and cross-border data flow provisions are established in contract; regulatory compliance (e.g., with GDPR) is embedded both contractually and technically.

Trust, social acceptance, and the removal of barriers for non-technical participants are recognized as key challenges, with efforts directed toward building transparent, intuitive dashboards and automating policy enforcement and usage auditing.

6. Open Research Challenges and Future Directions

Three principal research arenas structure outstanding problems (Perera et al., 2013):

1. Technological:
   • Achieving scalable brokered architectures spanning billions of sensors and high-frequency streams.
   • Developing zero-touch provisioning, ensuring plug-and-play heterogeneity across device classes and protocols.
   • Facilitating energy-aware sampling, efficient local fusion, and adaptive communication scheduling.
   • Advancing cross-platform interoperability with standard schemas, APIs, and semantic models.
   • Embedding security and privacy appliances at the edge.
2. Economical:
   • Lowering entry barriers for startups assuming SP/ESP roles, enabling ecosystem diversity.
   • Designing sustainable, transparent, and equitable financial models aligning owner incentives with consumer value.
   • Establishing data credibility and reliability mechanisms for mission-critical decisions.
3. Social and Regulatory:
   • Driving social acceptance by supporting change management and end-user education.
   • Harmonizing open data policies with emergent privacy laws and sector-specific regulations.
   • Automating compliance enforcement, particularly for usage auditing and misuse prevention.
   • Simplifying the user experience for non-domain experts.

The realization of SenaaS as a ubiquitous sensing utility for smart cities and beyond hinges on progress across these tightly coupled axes.

---

References:

Sensing as a Service Model for Smart Cities Supported by Internet of Things (Perera et al., 2013)