Keywords

1 Introduction

In current century, a large number of geospatial applications of various industries are running on edge devices like mobile, smartphones, tab with the help of modern technologies like Information and Communication Technology (ICT), Wireless Sensor Network (WSN), and Internet of Things (IoT). Different computing paradigms are involved to compute a large amount of geospatial data, communicate and provide meaningful geospatial information to the user. These geospatial applications help users to analyse trajectories, weather prediction, current traffic conditions, nearby restaurants, hotels, bus-stop, railway stations, ATM searches, etc. Using the long distant cloud server can be a major challenge for latency-critical geospatial applications. For example, a user in a moving vehicle wants to know the traffic conditions of the upcoming road junctions. The traffic movement information should reach the user within a certain time period. After reaching the road junction, no use in getting the information about the traffic condition of that particular road junction to the user.

Mobile Edge Computing (MEC) is a modern technology, where idle edge nodes of the network are used for data computation and storage purposes. It reduces the latency due to its presence near the user applications which is the main drawback of the remote cloud servers. This technology is a combination of mobile computing and wireless communications. If we consider the above traffic movement application. MEC will help by offloading computation-intensive tasks like vehicle counting by video cameras, traffic composition, different algorithms for traffic counting to the MEC servers for cloud execution.

Edge nodes of any network are varied in battery sizes. This battery is only a source of energy. While the applications are running in the edge nodes, the battery drains of nodes. It is required to recharge the battery of the nodes from time to time. Using green energy like solar, wind leads to reducing carbon emission by minimizing the burning of coal-based brown energy. Green energy is very much dependent on the weather and geolocation of the edge nodes. A combination of green and brown energy provides a continuous power supply to the edge nodes. Saving brown energy in the MEC network can be termed as Green MEC.

Many existing methods are available in the MEC domain to provide energy-efficient solutions. In the context of geospatial applications, all of the existing solutions are may not be suitable or applicable.

Several existing energy-efficient and delay-aware solutions can be utilized for geospatial applications.

In this chapter, we are investigating:

  • Existing geospatial applications used in MEC.

  • Challenges in the geospatial applications while running on the MEC.

  • Existing energy-efficient methods/techniques are available in MEC.

  • Future scopes of Geospatial MEC.

We discuss in the chapter possible challenges in Green MEC domains for various geospatial applications. Future directions of geospatial applications with green MEC will be discussed at the end of this chapter.

2 Mobile Computing Paradigms

Mobile Computing is a technology that enables users to communicate with each other wirelessly through messages, audio, and videos. Mobile computing is supported by two computing paradigms: (i) Cloud-based, and (ii) Edge-based.

  • Mobile Cloud Computing(MCC): Mobile applications are running on mobile devices (like smartphones, tab) but the data storage and processing have been done in the cloud servers. It enriches user experiences by executing a plethora of mobile applications on a large number of mobile devices. Sanaei et al. [1] defined the MCC as “a rich mobile computing technology that leverages unified elastic resources of varied clouds and network technologies toward unrestricted functionality, storage, and mobility to serve a multitude of mobile devices anywhere, anytime through the channel of Ethernet or Internet regardless of heterogeneous environments and platforms based on the pay-as-you-use principle.” The detailed MEC architecture is presented in Fig. 1. Public, Private, and Hybrid Cloud can involve this MCC architecture. Public Cloud is used for a wide range of users, whereas Private Cloud is used for an organization, institutions, or government usage purpose where privacy has utmost priority. Hybrid Cloud is used for flexible, cost-effective, agile, and scalable business purposes. It allows organizations to get the best of both private and public cloud models.

    Fig. 1
    figure 1

    Mobile Cloud Computing Architecture

    Full size image

  • Mobile Edge Computing(MEC): The cloud computing services are brought to the edge of the network by MEC. This technology is a key enabler for computation-intensive and latency-critical mobile applications [2]. Data storage and data processing are performed in the edge devices(e.g., base stations and access points) which are present in between mobile users and cloud servers. The main aim of MEC is to provide a better quality of service(QoS) to the mobile user by minimizing network congestion, and latency [3]. The detailed MEC architecture is presented in Fig. 2.

    Fig. 2
    figure 2

    Mobile Edge Computing Architecture

    Full size image

3 Existing Geospatial Applications on Mobile Edge Computing

A large number of geospatial applications are developed in the current decade which is running in MEC. We have categorized them into seven parts which are shown in Fig. 3.

Fig. 3
figure 3

Taxonomy of Geospatial Applications on MEC

Full size image

3.1 Smart City Services

Smart city is a concept where the quality of living of people is improved with the help of information and communication technology [4, 5]. Smart cities will improve various smart things like smart economy, smart living, smart environment, smart mobility, smart governance, and smart people [6]. Green communications are made through device-to-device communications in [7]. Also, spatial parameters’ effects on the smart city development and strategies are changed accordingly [8]. Here, we will discuss some smart city services where MEC plays an important role.

3.1.1 Traffic Prediction and Road Safety

Traffic prediction leads to an intelligent transportation system. It enhances the road safety of people. Traffic prediction can be done in several ways, i.e., traffic status prediction, traffic flow prediction, travel demand prediction [9]. Traffic predictions are done through a large number of Spatio-temporal data processing and techniques like data map-matching, data cleaning, data storage, and data compression. MEC can efficiently handle local/nearby Spatio-temporal data processing and provide local/region-specific traffic prediction.

Gillam et al. [10] proposed on-vehicle and off-vehicle data computation techniques in MEC. It increases autonomous driving efficiency by reducing end-to-end latency.

A machine learning approach was applied to Spatio-temporal data processing to get the mobility pattern of the moving agent in [11, 12]. It predicts the future locations of the moving agent in time-critical applications.

Emergency vehicles, like ambulances, can pass through quickly and safely on road through the safety corridors. A Back-Situation Awareness (BSA) application has been proposed by Halili et al. [13]. It notifies other vehicles about the presence or arrival of emergency vehicles. This helps to build a safety corridor for emergency vehicles. The nearby vehicles information is gathered through Multi-Access Edge Computing (MEC) with 5G.

Yang et al. [14] described short-term traffic prediction using the combination of vehicle velocity and traffic light prediction model. Vehicle velocity prediction has been performed with Spatio-temporal data analysis using a low-complexity semi-parametric prediction model. The model also captures traffic-light effects and driver’s behavior to identify real-time traffic changes.

The distribution of MEC resources is depending on the task uploading of applications. Ale et al. [15] proposed a Spatio-temporal based learning strategy for resource management. Resource distributions using Bayesian hierarchical learning methods provided better results than equal resource distribution over all servers. Wang et al. [16] proposed unmanned aerial vehicle (UAV) for flexible MEC services. UAV provided latency-aware services after merging the geographically nearby user requests.

Ridhawi et al. [17] described a MEC collaborative reliable and secure communications among IoT devices, edge layers, and cloud-fog layers. Traffic flows are reduced by using node-to-node communications. These node communications ensure reliable and secure communications. This methodology is useful for smart city services.

3.1.2 Health Care Service

Health care service is an essential service in smart city. Geolocation-aware health care service facilitates health monitoring near to the patients [18]. Medical facilities reach quickly from nearby health care centers after determining the geolocation of the patient. MEC helps to continuous health monitoring for indoor and outdoor patients in [19]. Outdoor patient’s mobility data prediction and preliminary health data analysis are done through edge devices. Any serious patient data analysis has been performed in a cloud server.

A deep learning-based heart disease identifying application, Healthfog, has been proposed by Tuli et al. [20]. Integrated IoT-Edge-Cloud platform is the backbone of the proposed architecture which made the application delay-aware and energy efficient.

A real-time healthcare service provisioning with geospatial queries in a cloud-fog-edge integrated platform has been proposed in [21]. The framework is an energy and latency-aware. It leads to a green geospatial query resolution platform.

Identifying the hotspot zone of Malaria [22, 23], Dengue [24] using geospatial map and taking immediate or preventive action accordingly are some research aspects in heathcare service.

3.1.3 Environment Monitoring

One of the environment pollutant gas is Carbon Monoxide (CO). Excessive emission of CO also increase heat of earth surface. To monitor CO level, a fog computing-based application has been proposed by Nugroho et al. [25]. Determination of the distance from the CO emitted area is calculated using Krigging method and plotted on Google map using lat/lon information.

Air quality at low concentration levels has been checked by AirSensEUR in [26]. Ganga river management using overlay analysis in MIST computing has been performed by Barik et al. [27].

3.2 Disease Monitoring

Disease identification and monitoring are necessary for contagious diseases. It controls the spreading of the disease over a geographic region. An IoT-based eHealth monitoring system has been proposed by Feriani et al. [28]. It identifies the symptoms(e.g., fever, coughing, and fatigue) of COVID-19. A learning method helps to identify the disease using MEC hierarchical framework.

3.3 Disaster Monitoring

High resolution and low latency face recognition videos have been identified the affected people in disaster prone areas [29]. A data distribution in nearby edge devices reduces the chance of data loss. Storing of disaster prediction data in nearby telephone central offices (TCOs) is proposed by Tsubaki et al. [30]. Face recognition of missing people in disaster is demonstrated by Liu et al. [31]. Facial images are transferred and stored in cloud server for saving energy and network bandwidth. Satellite image and video processed in a fog-cloud environment [32] to identify disaster-prone area.

3.4 Tourism Monitoring

Tourism helps in the economical growth of a country or a place. Monitoring of this sector is very important for business. It also has a socio-cultural impact. Behavior or movement pattern of tourists changes at a location on a seasonal or temporal basis. Tourism monitoring helps: (i) improve understanding about the effects of tourism, (ii) identify the locations where improvement is required, (iii) identify the season-specific locations, (iv) enable location-specific competitive business.

Ghazal et al. [33] captured and stored panoramic videos for a trip of tourist spots. Spatio-temporal information is also attached with videos for visualizing the videos with a location map. MEC architecture has been used for providing high-bandwidth and low-latency video experiences to end-users. Zhou et al. [34] used edge nodes for capturing Geo-tagged Flickr images. Image mining has been performed to determine the exact tourist location. Big spatial tourism data are managed through the RHadoop platform in Cloud. Van Setten et al. [35] developed a mobile-based recommendation system for tourists. A tourist guide application for Cyprus is elaborated in [36].

3.5 Geospatial Data Collection and Query Processing

Geospatial data are collected through edge devices. Geospatial queries are placed through geospatial applications using smart devices, i.e., mobile phone, tab, smart watch. To generate the result of the geospatial queries, huge data are needed to process along with the OGC compliant geospatial services [37] (WPS, WFS, WCS, WMS). MEC helps in processing local geospatial query processing with less delay. Also, storing, and processing geospatial data in a hierarchical manner to resolve geospatial queries has been performed in [38, 39]. This methodology not only reduces the latency of query result but also reduces the energy consumption of mobile devices as data processing is performed in edge or cloud platforms.

Essential data for services are collected from IoT devices and processed in MEC [40]. A lossy compression technique has been adopted to reduce the routing message contents. A budget and deadline constraint geospatial query processing have been done in cloud platform in [41]. Virtual Machine allocations for geospatial query processing has been performed in [42] with learning technique [43].

4 Existing Energy Efficient Methods in Mobile Edge Computing

Wu et al. [44] proposed an energy-efficient dynamic task offloading algorithm that decided where (MCC/MEC/IoT) tasks would be offloaded. Also, a blockchain-based offloading scheme has been proposed for secure task offloading. Lyapunov optimization-based dynamic secure task offloading and resource allocation algorithm described in [45]. They tried to extend the overall energy efficiency of non-orthogonal multiple access (NOMA) assisted MEC network.

Cheng et al. [46] proposed an energy-efficient algorithm for optimizing both offloading approach and the wireless resource allocation technique. It minimizes the transmission energy for computing task offloading. A greedy-based offloading algorithm [47] is described to minimize the overall energy cost for multiplayer games using MEC. Another energy-efficient task-offloading in MEC for an ultra-dense network is discussed in [48]. They reduced IoT devices’ energy consumption within latency deadline optimizing devices offloading decisions.

Zhou et al. [49] discussed a reinforcement learning technique for computation offloading and resource allocation in a dynamic multi-user MEC system. The objective of this work is to reduce the long-term energy consumption of the entire MEC system. An unmanned aerial vehicle based MEC is elaborated in [50]. Objective of their work is reducing the hovering time, scheduling and resource allocation of the IoT tasks. Non-orthogonal multiple access (NOMA) based resource and energy allocation methodology in MEC has been discussed in [51, 52].

We have represented the existing energy-efficient methodologies in Table 1.

Table 1 Existing work for energy efficiency in mobile edge computing
Full size table

5 Challenges in Geospatial Mobile Edge Computing

There are several challenges are present in Geospatial MEC. These are-

  • Heterogeneity: Different heterogeneous geospatial data sources are involved while resolving any user request. Also, edge devices and servers may present in heterogeneous networks. Proper synchronization and offloading policies need to develop among the geospatial data sources and computing resources to maximize utilization.

  • Seamless Connectivity: In MEC, the mobile devices are moving around and changing the region of MEC servers. Seamless connections are very much needed while a mobile device changes one MEC server region to another MEC server region. Otherwise, mobile applications, like mobile games may get hampered and the quality of the service gets down.

  • Cost: The amount of cost increases with the number of edge nodes involve in the edge network. Also, communication cost, coordination cost among edge nodes, geospatial data cost are taken into the cost calculation. CPU cost and IO costs are also involved for geospatial queries [41] processing in MEC. Accumulation of all these costs is needed to consider for geospatial MEC applications.

  • Data Handling: Geospatial applications are work with a large number of data/big data. MEC servers and MEC devices need to handle this huge number of data carefully. Otherwise, data may be lost during data transmission. Also, at the time of data analysis resource constraints arise due to improper data handling.

  • Security and Privacy: These two features are very important for critical mobile applications. Privacy of data may reveal during offloading. Also, security threats arise by masquerade attacks, passive attacks.

6 Future Directions

In this section of the chapter, we discuss the future scopes of the geospatial green mobile edge computing (GGMEC) research work. Though many explorations have been done in MEC and Green Computing, very little progress observed in the geospatial domain with energy efficiency. Still, we can think about the following aspects of Geospatial Green MEC in the future.

  • Pricing Scheme: Investigation of pricing policies is required individually for geospatial data providers and Mobile Edge Computing service providers.

  • Data Management: Geospatial data management in the GGMEC environment is a challenge. Storing a small amount of data within the edge nodes of a distributed manner and synchronizing them will be a challenging task.

  • Data Security: Sensitive geospatial data or healthcare data are required proper security while data are transmitted within edge devices.

  • Data Re-usability: Geospatial data can be reused for different applications. Storing of processed geospatial data is needed for reusing in future geospatial applications.

  • Energy Efficiency: A very small amount of investigation has been performed for energy efficiency in the geospatial domain. Large geospatial data processing and transmission lead to huge energy consumption in MEC nodes. Proper computing resource management and geospatial data management can provide Green MEC.

  • Application Management: Geospatial application management, MEC resource provisioning, with different kinds of machine learning techniques can be a future trend.

  • Application-aware Policy: Every geospatial application, i.e., smart city, weather prediction, disease monitoring, crop analysis, etc. has different requirements. Application-driven policies are required for proper management in the MEC environment.

  • Web Service Orchestration: Automation of different geospatial web service orchestration to resolve any geospatial query in MEC can also be a future scope.

7 Summary

In this chapter, we have discussed various geospatial applications in MEC. Also, we have provided a taxonomy for the applications. Different existing energy-efficient schemes in MEC are discussed and represented in a table. Thereafter, challenges in geospatial MEC are identified and discussed. At the end of the chapter, we have pointed out areas of geospatial MEC where further research can be explored.