Potential Transmission Choice for Internet of Things(IoT):Wireless and Batteryless Communications and Open Problems

Zhan Xu,Guanjie Hu,Minzheng Jia,Lan Dong

1 School of Information and Communication Engineering and with Key Laboratory of Modern Measurement&Control Technology,Ministry of Education,Beijing Information Science&Technology University,China

2 Beijing Key Lab of Transportation Data Analysis and Mining,School of Computer and Information Technology,Beijing Jiaotong University,Beijing,China

3 Department of Information Engineering,Beijing Polytechnic College,Beijing,China

Abstract:The 5th generation mobile communications aims at connecting everything and future Internet of Things (IoT) will get everything smartly connected.To realize it,there exist many challenges.One key challenge is the battery problem for small devices,such as sensors or tags.Batteryless backscatter,also referred to as or battery-free backscatter,is a new potential technology to address this problem.One early and typical type of batteryless backscatter is ambient backscatter.Generally,batteryless backscatter utilizes environmental wireless signals to enable battery-free devices to communicate with each other.These devices first harvest energy from ambient wireless signals and then backscatter these signals so as to transmit their own information.This paper reviews the current studies about batteryless backscatter,including various backscatter schemes and theoretical works,and then introduces open problems for future research.

Keywords:batteryless backscatter;battery-free;channel state information (CSI);channel estimation;multiple antennas;signal detection;symbiotic communication

I.INTRODUCTION

The Internet of Things(IoT)[1,2]has been attracting considerable interest in recent years [3].Future IoT aims at smartly connecting everything.As predicted by the forecast by Gartner,a world-famous leading research and advisory company,the number of IoT devices is growing exponentially and is expected to reach 26 billion by 2020[4].

According to the requirements of the 5th and 6th generation mobile networks [3],future IoT will connect everything intelligently,i.e.,enable physical objects to interact and exchange data with each other.Moreover,the 6th generation mobile networks propose the goal of smartly connecting everything.The objects to be connected include mobile phones,computers,and small computing devices such as radio frequency identification (RFID) tags,sensors,small computing units,and actuators to perform different functionalities.

Despite the ambitious aims,the development and deployment of IoT currently face many technical,economical,and even political challenges.As far as technologies is concerned,at least the following challenges exist:energy source for sensors,cost of sensors,algorithms or technologies of massive access,inter-connection architecture and internetworking protocols[5].

Let us focus on the energy challenge for IoT.The batteries for these small IoT devices,especially sensors and tags,have two disadvantages.That is,they have limited service lifetime and additional maintenance costs.When the batteries consumes all the energy stored,these small devices need new batteries and replacement.

Besides,the sensors may be limited by severe external conditions.For example,it is almost impossible to replace batteries of sensors embedded in walls.For another instance,sensors as well as their batteries along the beach suffer from sea water corrosion.

An efficient strategy to address the energy problem is to harvest energy from ambient environment,instead of utilizing time-limited and replacementrequired batteries.Existing possible candidates are solar,wind,vibrational and electromagnetic energy.These environmental energy sources have a series of applications.

However,solar,wind and vibrational energy are,in many cases,unstable and subjected to changes in the environment.For example,solar energy cannot be available at nights and wind energy may fluctuate with wind speeds.

In contrast,wireless signals of base stations (BSs),WiFi signals and radio television (TV) broadcast signals are pervasive and ubiquitous.Therefore,electromagnetic energy is a good choice for stable energy source for sensors in IoT.

Batterless backscatter,also referred to as batter-free backscatter,is a new technology that utilizes the wireless signals for energy source and also communication source.As shown in Figure1,small devices such as tags or sensors receive signals from base stations or RF sources,obtain energy from these signals,and then backscatter these signals for communication.In brevity,batterless backscatter can enable small devices to obtain energy from RF signal sources and to communication with communication nodes in networks.

The earliest type of batteryless backscatter is ambient backscatter proposed in[6].Later a series of types of batterless backscatter are suggested to enlarge communication ranges or enhance reliability

In this paper,we first show the difference between the traditional backscatter model and new batterless backscatter models.Next we summarize various types of batteryless backscatter and provide an overview of theoretical analysis about batteryless backscatter.Moreover,we point out the challenges for battery-free backscatter.

Figure1.Batterless backscatter systems.

The rest of this work is organized as follows.Section II introduces the difference between traditional backscatter and new batterless backscatter.Section III presents various schemes of batteryless backscatter.Section IV summarizes current theoretical analysis for battery-free backscatter.Section V provides open problems for batteryless backscatter communications in future IoT.Finally,Section VI concludes the paper.

II.TRADITIONAL BACKSCATTER AND NEW RISING BATTERYLESS BACKSCATTER

One successive and famous application of electromagnetic energy is radio frequency identification (RFID)systems,which represent the traditional backscatter.Traditional backscatter is a type of wireless communication that originated from 1948[7].It takes advantage of the reflection of the electromagnetic waves.

A typical RFID system contains a reader and a tag as shown in Figure2(a).The reader first generates a continuous electromagnetic wave.The wave will be received and then backscattered by the tag.The tag can add its own information during the process of backscattering.This process is often referred to as radio backscatter[7,8].

The basic principle of radio backscatter is that the tag responds to the reader by changing its antenna impedance and modulating its own information onto the backscattered wave [9].Since tags utilize exiting RF signals transmitted by the reader and thus do not need to generate these RF signals,radio backscatter saves the tag battery,and meanwhile lowers the costs of radio frequency(RF)components,and can also reduce the labor maintenance of replacement of batteries[10].

One shortcoming of conventional radio backscatter is the limited communication ranges between the reader and the tag due to a round-trip path loss.To address this disadvantage,bistatic backscatter,depicted in Figure2(b),is proposed in[11].Bistatic backscatter places an RF source,e.g.,a carrier generator,close to the tag so as to reduce the path loss between the tag and the RF source.

Another shortcoming of conventional radio backscatter is the requirement of special RF source.One dedicated power source (e.g.,a RFID reader) is required to drive the battery-free tag.To overcome this disadvantage,ambient backscatter,the earliest type of batteryless backscatter,is proposed in [6,12]and arouses a lot of research interests.

Ambient backscatter can enable small battery-free devices to communicate with other devices.It utilizes ambient RF signals,such as existing television(TV)and cellular signals,for energy source.It also remodulates and backscatters these signals for communications.Specifically,the small devices,i.e.,tags or sensors,will first harvest power from wireless signals,encode its information into binary bits,and transmit these bits by changing the antenna impedance.

The fundamental principle of ambient backscatter is that the tag can change its antenna impedance so that the incoming signals can be backscattered or absorbed.The backscattering or absorbing states can represent the binary bit:‘1’ or ‘0’.It is worth noting that later studies show that negative and positive backscatter can also be realized through changing the antenna impedance.Multiple backscattering states for batteryless backscatter will be a good or even better choices for high data rate transmission.

In summary,ambient backscatter,the first and typical type of battery-free backscatter,exploits wireless signals for both power and communication.Based on this,various types of battery-free backscatter are proposed and attracts much attentions from both academic and industrial circles since it can free small devices,such as tags or sensors,from life-limited batteries and dedicated energy sources.Judging from these,battery-free backscatter can be a potential technology for future IoT.

III.VARIOUS BACKSCATTER SCHEMES

A series of hardware prototypes based on batteryless backscatter technology has been realized.Figure2 lists several backscatter models that have been developed.For comparison,traditional RFID systems are also provided in Figure2(a).Besides,bistatic backscatter is depicted in Figure2(b).

The work in[6]demonstrates the feasibility of ambient backscatter communication with strong TV signals for the first time in 2013 (Figure2(c)).In addition,using existing Wireless Fidelity (Wi-Fi) signals,the authors in [13]suggest Wi-Fi backscatter with multiple receiving antennas and enable small RF-powered devices to have access to the Internet.Moreover,passive Wi-Fi backscatter is introduced in[14]to directly generate Wi-Fi transmissions through backscattering.

Since frequency modulation(FM)radio signals are pervasive on the earth,FM backscatter is presented in in[15].It is shown that FM broadcasting radio signals can power the tags and can help runners or passengers on travel to obtain local information through backscattering.

Miller writes that in dreams, old and dilapidated houses denote failure in business or any effort, and declining health (297).Return to place in story.

Full-duplex backscatter,described in Figure2(d),is presented in [16]to achieve simultaneous communication with both mobile uses and sensors for the WiFi gateway.The gateway is required to have two antennas,one for transmission a and the other for receiving.

Reference[17]proposes inter-technology backscatter described in Figure2(e).Inter-technology backscatter can transform Bluetooth signals to Wi-Fi and ZigBee-compatible signals.

One disadvantage of betteryless backscatter is limited communication range.To solve this problem,reference [18]develop a multiple-antenna cancelation and coding mechanism to enlarge the communication distance from 4 meters to 20 meters.Furthermore,reference [19]proposes LoRa backscatter where the communication ranges can reach more than 100 meters through spread spectrum technology.The LoRa backscatter system is shown in Figure2(f).

In the recently two years,large intelligent surface(LIS),also referred to as reconfigurable intelligent surfaces,are suggested to backscatter signals so as to provide extra paths to aid transmission from the transmitter to the receiver [20].LIS can be considered as an evolution of backscatter technology from one unit to massive units,which is becoming a hot topic.

Many other practical backscatter platforms are also suggested in recent years [21–24].These works can overcome the heterogeneity of the physical layer and realize cross-technology communication design between various networks including Bluetooth and Wi-Fi or ZigBee.

IV.THEORETICAL STUDIES FORBACKSCATTER COMMUNICATIONS

Batteryless backscatter is a new arising technology and most current researches focus on its early type:ambient backscatter.Theoretical studies of ambient backscatter have also attracted considerable attention,including coding schemes[25,26],channel estimation[27],signal detection[28–30],tag selection[31],and performance analysis such as bit error rate(BER)and channel capacity[32–34].

4.1 Coding

Reference [28],the first journal of theoretical analysis for batteryless backscatter,suggests a differential encoding scheme that can eliminate the necessity of channel estimation and can decode the tag information via energy accumulation,i.e.,signal power difference.Reference[25]designs a new coding scheme for tags with three backscatter states,unlike two states often assumed,to enlarge the throughputs of ambient backscatter systems.

4.2 Channel Estimation

Accurate channel estimation is of vital importance for wireless communication systems [35].An ambient backscatter communication system will perform better with better channel estimates.The channel estimation problems for ambient backscatter have been studied in[27]and[36]for single antenna and multiple antenna cases at the reader,respectively.

4.3 Signal Detection

Different from [28],reference [29]proposes a noncoherent signal detector for ambient backscatter communication networks without channel information(CSI)and also training symbols.Additionally,a semicoherent detector is provided in [37],where some composite parameters,based on the selection principle suggested in [35],are chosen for estimation and detection.These parameters are particularly selected from those related with key channel state information,and are estimated through using a few training symbols and unknown data symbols.

4.4 Multiple Antennas and Multiple Tags

4.5 OFDM Transmission

Ambient backscatter systems over orthogonal frequency division multiplexing (OFDM) signals is analyzed in [33]in terms of achievable transmission throughput and outage capacity.Reference [34]also suggests a method to exploit the cyclic prefix of OFDM so as to improve BER performance and achieve higher data rates.

V.CHALLENGES AND OPEN PROBLEMS

In this section,we provide open problems for future backscatter communications and practical challenges for its applications in future IoT.

5.1 Interference Analysis and Optimization

Batteryless backscatter utilizes surrounding RF signals for energy sources and backscatter them for communication.These backscattered signals may be received by other radio devices and thus result in interference.How to control and utilize these backscattered signals is an interesting topic.It is worth noting that symbiotic radio (SR) [38],a new paradigm,has been proposed to address this problem.It integrates backscatter device with a primary communication system,and the transmitters and receivers are jointly designed to optimize both the primary and backscatter transmissions.

5.2 Capacity or Throughput Analysis

Backscattered signals can be treated as extra path for the wireless systems where RF sources come from.Therefore,the original systems can benefit from the batteryless backscatter systems.As far as this in concerned,batteryless backscatter can be considered as a type of symbiotic communication.

One interesting problem arise as shown in Figure3.The original system that contain RF source transmit signals to the mobile user.These signals can power and motivate the batteryless backscatter system.Meanwhile,the mobile user can have extra paths from the backscattered signals.Therefore,we have two systems with the same RF signals:the original system and the batteryless backscatter system.Suppose the capacity of the original system isC0if the batteryless backscatter system does not work.However,if the batteryless backscatter system works,it will have new channel capacityCb;meanwhile the the capacity of the original system will increase fromC0to a new value defined asC1due to extra path from backscattering.Clearly,we may have

if the extra paths are fully exploited.The capacity analysis forC0,C1,andCbas well as the relationship between them can be a good topic for further investigation.

Figure3.Two systems:the original system and the batteryless backscatter system.

5.3 Multiple Antennas

Almost all current studies about ambient backscatter suppose that the tag is equipped with only one antenna[28,29,37,30,33,31,25,32].This assumption is reasonable because of simple design for small devices.

It is also worth noting that,in many practical cases,multiple-antenna tags can be advantageous for ambient backscatter communication systems.For example,if the tag has two antennas,it can harvest energy and transmit simultaneously.That is,a multiple-antenna tag can perform backscattering in one antenna and harvest energy in other antennas.This will increase reliability for batteryless backscatter systems because a single-antenna tag can choose only one states[6].

Another benefit of increasing the number of tag antennas is enlarged communication distance between the tag and the reader[39].Typical defect for batteryless backscatter systems is short transmission distance.Multiple antennas at the tag can increase diversity gain and overcome the distance limit to some extent[40].

More importantly,utilizing multiple antennas at the tag can facilitate signal detection.It has been pointed out in [41]that the reader can recover the tag signals without the knowledge of noise variance,RF source signal power,and even CSI if the tag can transmit with multiple antennas.This is a specially good choice,or tradeoff for batteryless backscatter systems where there is few training symbols available.

There exist many open problems when the tags are equipped with multiple antennas [41],such as energy harvesting and utilization,encoding and decoding methods [25],channel parameter estimation[35],antenna selection at the tag,signal detection at the reader [28],interference avoidance and cancelation [42],tradeoff between data rate and time delay,full-duplex division method[43],and optimization of wireless resources[10].

5.4 Security

Security problem is born with wireless communication systems[44].It is even more applicable to batteryless backscatter communication systems.It is not only because that the wireless channel are open,i.e.,accessible by any devices closed to the transmission source,but also that the backscattered signals do not belong to the batteryless backscatter communication systems.Therefore,it is of vital importance for backscatter communication systems to take special measures to guarantee its secure transmission.

5.5 LIS-related Transmission Technologies

Figure4.LIS-based backscatter communication systems.

Future backscatter communication systems can adopt LIS[45],which is similar to the evolution from single input and single output(SISO)systems to multiple input and multiple output (MIMO) systems and later massive MIMO systems [46].One feature for LIS-based backscatter communication systems is a number of channel parameters.As depicted in Figure4,there existsM ×Nfading channel parameters between the RF source and the LIS.Besides,there are alsoM ×Nfading channel parameters from the LIS to one antenna of the reader.If the reader is equipped with three receiving antennas,the number of channels will be tripled.Channel estimation will be extremely difficult for LIS-based systems;and adjusting these backscattering units in LIS is another issue that worths studies.

Unlike traditional backscatter communications such as RFID tags,LIS is used to promote or enhance existing communication links,rather than to send any information of its own.Besides,the array architecture(passive and active)and operating mechanisms(reflection and transmission) of LIS are different from these of large-scale MIMO systems.These features give rise to many new open problems for LIS-based systems.

5.6 Synchronization

Figure5.Backscatter aided transmission on railways.

Almost all theoretical studies about battery-free backscatter assumes perfect synchronization between the batteryless tags/sensors and the RF sources.In practice,this assumption does not hold.The small batteryless devices do not have powerful computing resources and also can only have limited or few training symbols for synchronization.Therefore,simple but reliable synchronization protocols or methods are needed for battery-free backscatter schemes in future IoT.

5.7 Backscatter Aided Transmission on Railways

Wireless transmission with high data rates on high speed trains is one key goal for future 5th and 6th mobile networks [47].Various scenarios,such as viaducts and tunnels,penetration loss at the train body of the wireless signals,and large Doppler shifts resulted from fast speed(360 km/hour)can lead to fluctuation in wireless channels and thus poor transmission performance.Traditional way to address this problem is relay.

Backscatter technology,instead of or together with relay,can also be utilized to overcome the timevarying channel fadings and improve BER performance.As shown in 5,the base stations transmit signals to the train antenna and the mobile user inside the train.Meanwhile,two LIS can cooperate and backscatter these signals.Many open problems about transceiver design exist in this scenario,including channel estimation and backscattering schemes at the LIS,beamforming between LIS and base stations,and signal detection at the mobile user.

VI.CONCLUSION

In this paper,we investigated the batteryless backscatter communication systems.Specifically,we presented its difference with traditional backscatter,concluded various types of backscatter schemes,summarized the current theoretical studies about battery-free backscatter,and provided open problems for its applications in future IoT.Solving these challenges will pave the way for extensive utilization of batteryless backscatter in the 5th and 6th mobile communication networks.

ACKNOWLEDGEMENT

This paper is funded by Scientific Research Program of Beijing Municipal Commission of Education No.KM201910853003.