A Proposed IoT-Enabled Smart Waste Bin Management System and Efficient Route Selection

,


Introduction
e revolution in communication and drastic computer application are the consequences of advancement in the Internet. From the traditional message transfer, today's era is of Web, machine-to-machine (M2M) communications, the communication of real-time audio, video, and sensory data, etc., which are possible due to advancements in super-media. Internet of ings (IoT) links the physical things such as vehicles, refrigerators, thermostats, and vehicles to the Internet over the wireless network and activate them to share and coordinate information [1][2][3][4][5][6]. e diverse applications of IoT include transportation, healthcare, manufacturing, home automation, and the power grid. IoT plays a remarkable role in improving the quality of life and growing the world's economy [7][8][9].
e Information and Communication Technology (ICT) developments generate more and more objects which are becoming embedded with the sensors and having the ability to communicate with other things and objects, that is, transforming the physical world itself into an information and knowledge system [10,11]. e IoT enables the things/objects in the environment to actively participant using internet. at is, they share data and information with other members or stack holders through wireless/wired networks, often using the same internet protocol (IP) that connects the Internet [12,13]. In this way, the things/objects are capable of recognizing actions and deviations in their surroundings and are capable of acting/reacting autonomously largely without human intervention in an appropriate way. ese future IoT growths need to see acceleration and a maturation of common standards, creative approaches, and more cross-sector collaboration with business models [14].
Wireless sensors are one of the important applications of IoT [7,15]. Many sensors may join to establish the network, namely, wireless sensor network, which is a self-organizing network. e sensors gather the sensed data back to the information center to be stored and analyzed in real time [16,17]. e sensors gather the data from their surroundings and send the data to the base station through intermediate nodes. ese sensor nodes can be deployed in diverse IoT applications as mentioned above for specific tasks. Moreover, the smart city application has attained high attraction in the last few years.
It has been investigated that efficient waste management is still missing in state-of-the-art research on smart city applications. erefore, an application, namely, Smart Waste Management (SWM) System is intended to be developed to overcome the garbage collection issues in smart city. e SWM provides on-time garbage collection and minimizes the total cost of the garbage collection process.
is work demonstrates that the smart waste management system in IoT empowers the cleaning operators to detect cleaning issues in real time. erefore, this system ensures increase in the overall productivity and cleanness. Moreover, the end-to-end delay is also taken into consideration for the proposed architecture. e term "end-to-end delay" is demonstrated as the sum of the time taken by a packet to reach from the source to destination which is the resultant of many factors such as interference level along the path, the length of the routing path, and the number of nodes/hops in the routing path. is article focuses on the nodal delay, which is the delay from the source to destination and the delay at the single node.
Traditionally, waste bins are deflated at certain intervals by cleaners.
is procedure has numerous drawbacks as follows [18]: (1) Some garbage bins are filled up faster than the rate of picking up and emptying while they are filled up before the next time scheduled for collection which postures hygiene risks (2) ere are some special days, for example, public holidays, weekends, and festivals, during which many waste bins fill up very quickly that may require faster collection interval e work proposed here illustrates the smart waste management architecture that ensures the cleaning operators for detecting cleanliness issues in real time. Moreover, minimizing the end-to-end delay is also taken into consideration when a message is required to be sent for garbage collectors. e article is organized as follows. In Section 2, the background of the Internet of things is discussed. e proposed methodology is described in Section 3. Section 4 describes the simulation results. Conclusion and future works are described in Section 5.

Background of Internet of Things
Communication is the most vital part of IoT, that is, the interconnectivity of different devices which are able to communicate with other devices [19]. All the additional properties, such as maneuvering, sensing, capturing, process, and data storage, are unnecessary unless and until devices precisely require one of these properties. However, the communication ability is necessary when classifying a device as an IoT device. It is less important than the communication being performed; meanwhile, the physical and link layer communication within IoT can be comprehended in many ways. Figure 1 shows that the communication of devices does not always require a specific communication network. at is, when two devices are in communication range, they can directly communicate with each other; for example, radio technologies, such as Bluetooth or ZigBee, enable direct communication [20]. In Figure 1, devices communicate via a gateway using one protocol, that is, IPv6 over "Low-Power Wireless Personal Area Networks" and then the gateway could communicate using another protocol that is IPv4 which is the Internet. In Figure 1, case B demonstrates two devices are directly connected to the communication network, while not requiring a gateway, and are able to communicate even when nodes are positioned in various places. A physical device can be plotted in the information world by using more than one virtual thing; however, a virtual thing does not need to be associated with multiple physical things. It will have existence of any physical thing autonomously; for example, a physical thing can be executed by multiple applications having multiple characteristics in the virtual world. In the same way, the virtual things also have many identities; for example, a virtual thing may be an audio file in a USB drive. is audio file may have different file names and multiple copies, and these files may possibly have different qualities.
In [21], the transport protocol is introduced for efficient delivery of super-media data from smart surveillance systems. Flow control and synchronization techniques are taken into consideration to maximize the efficiency of smart surveillance systems. ere are three channels considered in the proposed architecture between the collected data from the environment and the data center. e first channel is the video channel, the second channel is the audio channel, and the third channel is the multisensory channel.
Learning-based synchronous (LS) approach is presented in [22] to minimize the transmission delay for Industrial Internet of ings (IIoTs), while showing no effect on reduction of network lifetime. LS algorithm considers forwarding nodes to minimize the transmission delay for IIoTs. In the LS scheme, each sender node looks the intermediate node which is closer to the sink node is selected through which the data are communicated to the sink node. e sender node synchronizes its duty cycle with its neighbor node by self-learning. In sensor-based IIoTs, the nodes near to the sink consume high energy than the nodes far away from the sink. However, the nodes nearer to the sink have lesser delay, which eventually enhances the network performance in terms of delay.
Utility-based adaptive duty cycle (UADC) is presented in [23] that shows robust energy efficiency, minimizes transmission delay, and increases the lifetime of the network. e algorithm considers comprehensive performance evaluation to determine the selection of relay nodes. In data dissemination, the algorithm selects the nodes which improve the efficiency in terms of energy conserving. e transmission efficiency is enhanced via maximizing the attempt of retransmission of energy consumption per unit transmission. To gain the superior value, the forwarding nodes consisting of small unit transmission bit energy dissipation, minor delay, and high energy are considered.
us, the overall energy efficiency is achieved, energy utilization is stabled, and delay in data dissemination is minimized. Accordingly, the overall UADC achieves an optimum outcome from the network via selecting a link that increases the network's function.
Waste management system (WMS) is proposed in [24] that focuses on the social aspects of the waste management systems. WMS uses IoT devices with RFID tags, weight, and ultrasonic sensors to measure the citizen interaction with the WMS. Each bin is equipped with an IoT device, ultrasonic sensor, weight sensor, RFID reader, solar cell, and two light emitting diodes as indicators, such that one of these indicators is colored in red and the other is colored in green. Every citizen who wants to interact with the system have an RFID card associated with a credential for the citizen. However, architecture is not efficient in terms of bin management, resource conservation, and message dissemination. e time taken by a message to travel from the source to destination is called end-to-end delay. e delay depends on the number of hops and congestion on the network. e sensor nodes in wireless sensor networks are powered by small batteries that can be charged or replaced. Hence, sensors can only send a finite number of bits from the source to sink until they run out of energy [25]. e major metrics for QoS is end-to-end delay. Both data rate and end-to-end latency are a combined effect for user-perceived data transfer speed. For the transmission of the small-sized file, end-to-end delay is the dominating factor, and for transmission of large-sized file, the data rate is the dominating factor. In wireless sensor environment, where the sensor nodes are required to be reported periodically to the sink, end-to-end delay plays a key role. For this purpose, we focus to find and minimize the endto-end delay in wireless sensor nodes. e term "end-to-end delay" is described as the sum of time taken by a packet to reach the destination. Many factors can be resultant of it, including the interference level along the path, the routing path length, and the number of hops in that routing path.
So far, our focus is on the nodal delay in smart waste management architecture, which is the delay from the source to destination and the delay at the single node. Now, let us observe the total number of delay from the source to destination. Suppose we have "N − 1" numbers of nodes from the source node to the destination node. Let us consider for a while that the network is uncongested so the "queuing delay" is insignificant, and at each router, the processing delay exists and host source is D proc. Each router transmission rate out of the source host is R bits per second while D prop is the propagation on every link; the delays at each node accumulate the end-to-end delay: where L is the packet size and R is the router.

Proposed Methodology
is section provides the proposed methodology which demonstrates the smart waste management system to empower the cleaning operations and to detect cleaning issues Journal of Computer Networks and Communications in real-time IoT. e Smart Waste Management Architecture is presented in Section 3.1 that ensures increase in overall productivity and cleanness. Section 3.2 presents endto-end delay minimization when a source node (bin) is full of garbage and needs to be emptied. Both the sections are described in the following.
3.1. Smart Waste Management Architecture. One of the difficult operational problems of municipal and local authorities are facing is the collection of municipal solid waste.
In recent years, due to environmental concerns and number of costs, most of the municipalities have been forced for assessing their solid waste management and examining their cost-effectiveness and environmental impact, for example, designing the collection of routes. During the past 15 years, numerous technological advancements, new acquisitions, and developments were provided in the industries. Consequently, both municipal and private haulers are giving serious considerations to use advanced technologies such as computerized vehicles that take the decision of route planning and schedule of collection of waste [26]. is proposed system describes a study of planned and computerized vehicle routes for the collection of municipal solid waste in the different regions of Pakistan.
e proposed system contains all the stages from the collection of our waste material, load to the truck, and recycling it in the recycling unit.
For this purpose, it includes the following list of activities: (i) Controlling and monitoring the collection of wastes; efficient and speedy transportation to the recycling units/point (ii) Preventing the waste from spilling from the waste bins during transportation to the recycling units (iii) Speedy transportation to recycling units so that the traffic condition will not be bothered in peak hours (iv) Proper storage and maintenance in the storage units is is the only one aspect of recycling of the waste. Now, let us talk about the smart waste container recycling and management system. We have used smart bins in which waste-detecting sensors are fixed. ese sensors are capable of sending signals to the nearest sensor referenced to the base station.
When the sensors detect the volume of the garbage, they communicate to the respective centers, indicating the volume of the garbage.
e garbage centers will have the garbage pickup truck sent to collect garbage from the garbage bin and recycle it in the recycle unit.
In the following diagram, it is indicated that the garbage pickup goes on the routes which are the most optimized routes, both cost-wise and hygiene-wise.
In Figure 2, it can be easily noticed that the pickup truck is directed to those waste bins which are more than 90% full or about to be 100% full and not selecting the path which is partially filled. In this case, the pickup truck selects the most optimized route that not only reduces its costs in the collection round trips but also shields from the unfavorable conditions.

e Proposed End-to-End Delay Technique.
e estimation and calculation of the delay estimator at any node n i are performed by using equation (2) which is given in the following, where TD i,j is the delay which is caused by transmitting the packet (P) to the node n j from the node n i using the link L i,j and QD ni is the queue delay. P r is the propagation delay from one node to another node while processing delay can be put in other delays which are experienced by the packet (p); however, it can be ignored as these are negligibly small: Queue delay QD ni of packet P is the delay at any node before processing of the packet P which remains in the queue. e smoothing factor value is constant α which is 0 and 1. Initially, QD ni is the queue delay experienced by the first packet: e transmission delay (TD i,j ) is the amount of time taken by the packet (P) which remains in the MAC layer of node n i before successful transmission to the node n j on link L i,j or dropped. It can be calculated by the formula given in equation (4), where the data rates in bits are represented by DR bits, the size of the packets are represented by SP bits, and the number of transmitted packets in time interval δt is NP: Propagation delay, P r D, is the amount of time taken by the packet to make physical journey from the node n i to node n j . d ni is the distance divided by velocity v as given in the following equation: If there are N paths from the sender node to the destination node, then the sender estimates the Path Delay PD and choses the path that has least delay comparative to other paths to send messages. Let n p denote the number of paths and n i denote the number of nodes, where I � 1, 2, 3, . . ., N.
Path delay (PD) on each path is estimated as follows: If p > 1, where p denotes the number of paths, then e transmission is considered from a source to a destination with the minimum end-to-end delay over the wireless network where transmission can be reserved and guaranteed on the links. Different paths will be required for different intervals of values that compute a table that maps all intervals for corresponding paths that minimize the endto-end delay as described in Figure 3. e responsibility of EREP algorithm is to select the desired next hop based on the demanded QoS parameters as described in Algorithm 1. Upon receiving the data packets (DPs) from the data packet classifier module, our proposed EREP algorithm, given in the following, searches the routing table (R t ) for only those nodes in the neighborhood whose link quality (L-QUAL i,j ) is greater than or equal to the predefined threshold level (L-QUALthre) and places them in NNL-QUAL (lines: 2-4). e DP is discarded immediately in case of empty NNL-QUAL (lines: [5][6]. Otherwise, in case of DP belonging to CD, delay aware procedure is called with inputs NNPD and DP (lines: [7][8]. While in case of DP belonging to NCD, the desired hop (DH) is the node belonging to NNL-QUAL.

Journal of Computer Networks and Communications
Once delay aware procedure is called, it selects only those nodes belonging to NNL-QUAL, whose end-to-end path delay (PD i,j ) is less than or equal to required delay (PD req ), and places them in NNPD (lines: [10][11][12]. In the case of empty NNPD, the data packet is discarded immediately (lines: [13][14]. If there is a single node in NNPD, then that node is selected as DH (lines: [15][16]. Otherwise, the reliability-aware procedure is called with inputs NNPD and DP (line: 17).
Upon calling the reliability-aware procedure, it selects only those nodes whose end-to-end path reliability (PR i,j ) is greater than or equal to the required reliability (PR req ) and stores them in NNPR (lines: [18][19][20]. If none of the nodes fulfills the required reliability demand, then the DH is the node belonging to NNPD having highest PR i,j (lines: [21][22]. If there is a single entry in NNPR, then that node is selected as DH (lines: [23][24]. Otherwise, the node belonging to NNPR with least PT i,j is selected as DH (line: 25).

Simulation Results
To study the performance of the proposed algorithms and to compare them to the existing ones, a wide experiment is conducted using a simulation tool. MATLAB is a highprecision level numerical-simulation environment and a popular tool to evaluate network models, which is used as the simulation tool. e simulation network is considered in Table 1 to evaluate the performance of the algorithm. e simulation results are analyzed to evaluate the efficiency of the proposed algorithms. Various performance parameters, like packet loss, congestion, packet delivery, ID Dest � Destination ID LOC dest � Co-ordinates of the NOD ID n,J � ID of the Neighbor Node n j L-QUAL i,j � Link Quality between nodes n i and n j PR i,j � Path Reliability from node n i to destination through node n j PD i,j � Path Delay from node n i to destination through node n j LOC n,j � Co-ordinates of the neighbor node n j Inputs: Rt and Data Packet P (1) for each data packet DP do (2) for each node n i ∈ RT do (3) if L-QUAL i,j > � L-QUAL then (4) store node n i into NNL-QUAL (5) if NNL-QUAL � � NULL then (6) discard DP immediately (7) else if DP ∈ CD (8) call Delay Aware Procedure (9) else DH � n j ∈ NNL-QUAL Delay Aware Procedure (10) for each node n i ∈ NNL-QUAL do (11) if PD i,j > � PD req then (12) store node n i into NNPD (13) if NNPD � � NULL then (14) discard DP immediately (15) else if NNPD � � 1 then (16) DH � n j ∈ NHPD (17) else call Reliability Aware Procedure Reliability Aware Procedure (18) for each node n i ∈ NNPD do (19) if PR i,j > � PR req then (20) store node n i into NNPR (21) if NNPR � � NULL then (22) DH � n j ∈ NNPD with highest PR i,j (23) else if NNPR � � 1 then (24) DH � n j ∈ NHPR (25) else DH � n j ∈ NNPR ALGORITHM 1: Efficient Route Election procedure (EREP).

Congestion.
e network congestion can be defined as reduction of quality of service (QoS) which occurs in those cases when a network carry more data than it can handle.
is affects the queuing delay, resulting in the blockage of connections and loss of packets.

Packet Delivery.
Packet delivery is the ratio of generated data packets to the ratio of data packets received at the destination.

Buffer Size.
Buffer size is the physical region of memory storage that is used by temporarily stored data which are being moved from one place to another.

Bit Rate.
Bit rate is the rate at which data are transferred from the source to destination. In other words, we can say that in a given amount of time how much data are transmitted. It is mostly measured by bits per second (bps) or (Kbps) and (Mbps).

Average Packet Loss Ratio.
e performance of the algorithm can be evaluated using the average packet loss ratio where a better algorithm can be defined when the packet loss ratio is least upon the data generation increase. Figure 4 describes the performance of efficient route election Journal of Computer Networks and Communications procedure (EREP) in terms of packet loss ratio, while the result for EREP is compared with CDR and LTRT procedures.

Average Success Ratio.
e performance of the algorithm can be evaluated using the average success ratio where a better algorithm can be defined when the packet success ratio is greater. Figure 5 describes the performance of EREP in terms of average success ratio, while the result for EREP is compared with CDR and LTRT procedures. e impact of data generation rates is shown in Figure 5 over packet success ratio by considering different link qualities. With the increase in data generation, the packet success ratio is decreasing slightly for all schemes which might be due to congestion.

On-Time Packet Delivery.
e performance of the algorithm can be evaluated using on-time packet delivery where a better algorithm can be defined when the on-time packet delivery is greater while the data generation is increasing. Figure 6 describes the performance of EREP in terms of on-time packet delivery, while the result for EREP is compared with CDR and LTRT procedures. e time constraint impact on the average time packet delivery ratio is similar by considering different link qualities.
It is clear from Figures 4-6 that the proposed EREP scheme outperforms than other current schemes such as CDR [27] and LTRT [28]. is is because of considering propagation delay during evaluation of end-to-end delay on each hop.

Conclusion and Future Work Direction
e article proposes IoT-enabled waste management system (SWM) for smart city applications. erefore, an application is developed initially in the smart cities, namely, the smart waste management (SWM) system.
e SWM system provides on-time garbage collection which eventually minimizes the total cost of the garbage collection process. e proposed work demonstrates that the waste management system in IOT empowers the cleaning operators to detect cleaning issues in real time. erefore, this system helps in increasing overall productivity and cleanness. Moreover, the end-to-end delay is also taken into consideration for the proposed architecture. e term "end-to-end delay" is demonstrated as the total time taken by a packet to reach from the source to destination which is resultant of many factors, including the interference level along the path, the length of the routing path, and the number of nodes/ hops in the routing path. Propagation delay is also considered during evaluation of end-to-end delay. It has been observed from the simulation results that the proposed scheme outperforms existing works.
Maintenance algorithm, reliability, redundancy minimization, and data broadcasting are open research issues in SWM application.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.

Authors' Contributions
Asim Zeb proposed the methodology. Qurban Ali was responsible for validation. Muhammad Qaiser Saleem reviewed the manuscript. Khalid Mahmood Awan was responsible for resource management and conceptualization. Jamal Uddin and Muhammad Qaiser Saleem were involved in funding acquisition. Ali Saeed Alowayr and Saleem Iqbal wrote the manuscript and were responsible for final proof approval of the manuscript. Faisal Bashir carried out reviewing and editing.  Journal of Computer Networks and Communications