Reconfigurable Intelligent Surfaces for 6G IoT Wireless Positioning: A Contemporary Survey

The sixth-generation (6G) wireless communication system is expected to integrate communication, intelligence, sensing, positioning, control, and calculation to adapt to time critical, ultrareliable, and energy-saving data delivery, as well as accurate positioning of personnel and equipment, serving the Internet of Things (IoT). On the one hand, reconfigurable intelligent surface (RIS) can intelligently manipulate radio waves and is considered to be one of the candidate technologies for the 6G wireless communication. Hence, there are more and more surveys on RIS-assisted communications. On the other hand, the potential of RIS in positioning has attracted growing attention, and articles on RIS-assisted positioning have been blown out. Therefore, it is time to review this literature to understand the potential of RIS positioning, research status, and point out the direction for future research. This article first explains the working principle and channel model of RIS and summarizes some characteristics of RIS suitable for positioning. Then, we give a concise review and classification of existing RIS positioning research. Finally, we put forward our views on the future research challenges and attractive directions for RIS-aided wireless positioning technology.


I. INTRODUCTION
W HILE 5G is currently being deployed around the world, efforts from industry and academia have started to look beyond it and conceptualize sixth-generation (6G) [1], [2], [3], [4], [5]. It is reported in [6] that global mobile traffic growing at an annual rate of around 55%, and this traffic is predicted to exceed 5 ZB/month in 2030 with the estimated number of global mobile subscriptions up to 13.8 billion in 2025 and 17.1 billion in 2030. The future large-scale industrial Internet of Things (IoT) and Internet of Vehicles (IoV) systems put forward new and higher requirements on communication systems in terms of delay, throughput, and reliability. Coverage enhancement (CE) is a major and challenging issue in IoT deployment. Security/privacy [7], [8], energy efficiency [9], and large-scale connectivity are other major issues. The application of the IoT, the corresponding requirements, and the limitations of 5G make it necessary to develop 6G [10], [11], [12].
Future mobile communication systems are expected to accomplish the vision of the Internet of Everything (IoE) with versatile networks designed not only for ubiquitous communications but also for accurate and reliable seamless positioning, which is an important foundation of the IoE. The increasing demand for location-based services, such as navigation, smart city, extended reality (XR) services [encompassing augmented, mixed, and virtual reality (AR/MR/VR)], healthcare monitoring, autonomous driving, and indoor positioning has led to a growing interest in ubiquitous positioning [5], [13], [14]. To fulfill this grandest vision, it is necessary to develop high-performance localization techniques, not only to improve the communication performance in various aspects at different network layers but also to meet the requirements of various emerging commercial and industrial on locationbased services. The 6G wireless communication system will not only provide communications between people, mobile devices, machines, and objects, but will also integrate communications, intelligence, sensing, localization, control, and computing to accommodate the time critical, ultrareliable, and energy-efficient delivery of data, and the accurate localization of people and devices [3], [4], [5], [11]. As shown in Fig. 1, 6G's requirements for positioning show six major trends. The positioning accuracy provided by GPS in the 4G system is about 10 m. 5G, using a much larger bandwidth and higher carrier frequencies, combined with antenna array deployment at a user equipment (UE) and base stations (BSs), is expected to further improve the location accuracy to 10 cm in 2-D [17]. Assuming that the maximum bandwidth for 6G will be up to 100 times that of 5G systems, its potential for positioning accuracy and efficiency will be much greater, so that it is envisioned that localization precision will be as high as 1 cm in 3-D [17], [18]. 5G has moved toward the millimeterwave band, and even 6G is expected to expand to the THz band. However, 6G THz high-frequency transmission is more sensitive to the obstruction of obstacles in the environment, the line-of-sight (LOS) transmission of the signal is easily blocked. NLOS propagation causes reflection, refraction, diffraction, and scattering of the positioning signal, resulting in a measurement error.
These phenomena are the main limiting factors for maximizing wireless network performance. Although a large number of studies have been used to mitigate the effects of multipath and NLOS, most of the existing studies are conducted under the assumption that the wireless environment is an uncontrollable factor. With the surge in demand for wireless networks, only making improvements at the sending and receiving ends of the wireless environment may not be enough to meet the challenging requirements of future wireless networks. To tackle these challenges, the concept of a smart radio environment (SRE) or, more recently, intelligent radio environment (IRE) is proposed [19]. The environment is regarded as an optimized variable, which can be programmed and controlled to make it play an active role in processing information.
Reconfigurable intelligent surfaces (RISs) are a key enabler to promote the realization of the SRE vision. RIS is a planar surface consisting of a massive number of reflecting elements, capable of reflecting the signal received from the transmitter to the intended receiver in a controlled manner, actively customizing the wireless environment. RIS combines a low profile, lightweight, and small geometry with low cost and high flexibility and they can be integrated into existing networks without changing hardware and standards but only the communication protocols [20], [21], [22]. For these reasons, they can be easily deployed anywhere, including stationary or moving objects and even on people. For example, RIS can be installed on the outside of buildings, street lamps, billboards, ceilings, people's clothes, vehicles, unmanned aerial vehicles (UAVs), planes, satellites, and so on, as visualized in Fig. 2. Therefore, RIS will have the opportunity to assist the 6G space-airground-underwater/sea integrated network in the future to further realize the vision of the IoT. It can be used for satellite communication applications, nonterrestrial network (UAV) communications, ultradense ground networks, underground and marine communications networks, for smart cities (smart homes, autonomous driving, smart factories, smart healthcare, etc.), industry 4.0, and positioning indoors and outdoors.
Presently, a massive MIMO technology in 5G communication systems is relatively mature. However, in the process of moving to 6G in the future, due to the use of THz frequency bands, MIMO requires dozens or even hundreds of antennas and radio links at the BS. For example, the use of terahertz in a 6G network will enable the number of transmitting and receiving antennas of the ultramassive MIMO (UM-MIMO) system to reach 1024 [23]. Large-scale deployment leads to high hardware costs and power consumption, as well as complex  Table I. The most similar to RIS is MIMO, but they are fundamentally different. In terms of usage, MIMO is based on the concept of transmission and reception, and a major feature of RIS is that it can play the role of a reflector. In terms of energy consumption, MIMO requires a large number of radio frequency (RF) links and needs to provide a large amount of additional power. Since RIS uses a large number of passive reflective elements and can be considered almost passive, energy consumption is greatly reduced. In addition, RIS itself does not have an RF link, no matter it is in the millimeter-wave frequency band or the terahertz frequency band, there is no need to worry about cost and power consumption. Therefore, with its advantages of cost saving and power consumption, RIS is a promising concept in the next-generation mobile communication system.
Although all these advantages make RIS an attractive technology for 6G in general to assist communications [20], sensing [24], simultaneous wireless information and power transfer (SWIPT) [25], [26], the possible applications and technical challenges of RIS-assisted localization need to be investigated [18], [27]. In terms of high-precision positioning, RIS can create virtual LOS links to deal with NLOS problems. And, it can intelligently manipulate the outgoing radio waves to focus the signal to the target user position. RIS can be controlled by a fast processing field-programmable gate array (FPGA) to meet the low latency requirements of 6G positioning. In addition, RIS has the potential to play an indispensable role in the diverse application scenarios of 6G networks and can assist 6G smart connections, ubiquitous connections, deep connections, and holographic connections with extremely low cost and power consumption. It can be concluded that the application of RIS to 6G positioning needs to be studied in depth. To this end, this article is devoted to reviewing the research progress on the use of RIS for localization, and the main challenges and opportunities are anticipated. This is not the first survey on RIS-assisted wireless localization, since there are other works on the subject, such as [18]. However, Wymeersch et al. [18] summarized the challenges and opportunities of RIS-assisted positioning from the perspective of radio positioning and RIS mapping. Compared with the survey, our  work discusses the research status and development potential of RIS-assisted positioning from a macro and long-term perspective.
The remainder of this article is organized as follows. In Section II, the hardware structure and channel model of RIS are first received. According to the characteristics of RIS, the potential of RIS for positioning is analyzed. In Section III, the latest research progress of RIS-assisted positioning is given. In Section IV, the challenges faced by RIS and potential future research directions for RIS-assisted wireless localization are outlined, followed by conclusions in Section V.

A. How RIS Works
The RIS is an artificial material composed of a 2-D array of metal and dielectric elements [14]. Arun and Balakrishnan [28] designed the first large-scale real-world RFocus system prototype with 3720 low-priced antennas on a 6-m 2 surface, as illustrated in Fig. 3. In 2020, NTT DOCOMO operators in Japan recently exhibited a metasurface-based RIS product (namely, artificial smart glass), which can dynamically control the response of impacting radio waves, including three modes of full penetration, partial reflection, and total reflection, as shown in Fig. 4 [29]. This smart glass is so transparent that it is very suitable for installation in buildings, indoors, vehicles, billboards, etc., without affecting people's sight and areas that require confidentiality. In order to understand how RIS works, the hardware structure of RIS is first introduced. As illustrated in Fig. 5, RIS has a three-layer structure and is equipped with an RIS controller. On the outer layer, numerous tunable elements (also called reflection elements/meta-atoms) are printed on the dielectric substrate to interact with the incident signal. The arrangement of these tunable elements is usually a repeating pattern. The difference between the reconfigurable metasurface and the metasurface is that the meta-atoms of the former can be dynamically configured, while the meta-atoms of the latter are fixed during the design and manufacturing stage. The size and thickness of the elements are between (λ/10) ∼ (λ/5) with a spacing of (λ/2) or less, where λ is the wavelength [30], so RIS is a thin surface. RIS supports full-band transmission as long as the dimensions of the RIS components are designed to match the operating frequency. In order to prevent signal leakage, a copper backplane is placed on the middle layer. The inner layer is a control network that is used to adjust the reflection coefficient of tunable elements. A very promising, cost effective, and highly scalable approach is to control the metasurface switches as a diode array [30], such as PIN diode or varactor diode array, as shown in Fig. 6. By setting the voltage, the switching state of the diode can be changed to control the state of the tunable elements on the outer layer. The inner layer is the link that connects the RIS controller and RIS units. Its function is to receive the voltage signal output by the RIS controller to control the RIS and pass it to the reflecting elements. The RIS controller is used to receive commands from the central controller or BS to adjust the amplitude and phase of the incident signal for each small reflecting element. Alternatively, the RIS controller can run independently using AI algorithms to adjust the state of RIS. In addition, an RIS controller, implemented by FPGA [21], also completes the connection with all peripheral modules, such as direct current (dc) power supply and timing module.
When an incident electromagnetic wave hits the surface of the RIS, a surface current will be generated. The current distribution is usually determined by the size and spacing of the RIS units and the different configuration states of the RIS inner control network. The control network generates different RIS configurations by adjusting the bias voltage of the PIN diode, which generates different surface currents, affects different radio wave responses, and realizes intelligent control of reflected signals. After the signal passes through the RIS surface, the reflection angle is no longer equal to the incident angle, which also proves that RIS has the ability to reflect the signal to the designated user location, breaking the law of reflection.

B. RIS Singal and Channel Model
When the RIS works in the reflection mode, each RIS reflection unit can independently reflect the incident electromagnetic wave signal. Assuming that RIS has N reflection units, n is recorded as the nth reflection unit, and n = 1, . . . , N. For each reflection unit, the baseband equivalent signal model can be written as follows: where x n and y n represent the incident signal and reflected signal, respectively, β n ∈ [0, 1] is the reconfigurable reflection amplitude of each unit, θ n ∈ [0, 2π) is the reconfigurable reflection phase, and β n e jθ n is called the reflection coefficient of the RIS reflection unit. Extending to the entire RIS surface, since each reflecting unit is independent of each other, the relationship between the signal reflected by the RIS and the incident signal has a diagonal matrix form, which can be written as follows: where is the reflection coefficient matrix of RIS. Note that each element of the RIS is designed to reflect the incident signal to the maximum so that β n usually assumes the value of 1. However, the value of β n in the real model is a constant related to the realization of a specific circuit, and the mismatch with the assumed model will affect the performance of near-field positioning, especially at a high signal-to-noise ratio (SNR) [33]. Although β n and θ n can be adjusted continuously within an interval, they are limited by cost and complexity in actual implementation, so discrete values are a design that can be considered. There are three types of reflection coefficients, including constant amplitude with continuous phase shift, optimized amplitude with continuous phase shift, and constant amplitude with discrete phase shift, which leads to three different constraints on the reflection coefficient in practice. In papers [20], [34], [35], [36], [37], [38], [39], the reflection coefficient is assumed to be a continuous phase shift. However, in practice, reflecting elements that use continuous phase shift are limited by hardware with high costs, especially when the reflecting elements of the RIS are very large. Therefore, in practice, the application of RIS usually uses discrete phase shift to significantly increase cost effectiveness [40]. However, it also has an inevitable problem that discrete phase shifts cause misalignment of RIS-reflected signals on the ideal receiver. Without loss of generality, we assume a positioning scenario where the multiantenna BS is located in the P B = [x B , y B , z B ] T and is equipped with N B antennas. RIS has N reflecting units, the coordinate of the center of the surface In the wireless positioning system, a large number of positioning devices are equipped with a single antenna (smartwatch and smart vest) and multiple antennas (robots and self-driving cars). We suppose mobile station (MS) The propagation path consists of a direct path and a reflection path through the RIS. The equivalent channel responses from the BS to the RIS, from RIS to the MS, and the BS to the MS are represented by The incident signal is set to x, the received signals y can be expressed as follows: The noise generally obeys the distribution of a complex Gaussian random variable with mean μ and variance σ 2 and can be expressed as n 0 ∼ N (μ, σ 2 ). In the process of actually establishing the channel model, the following factors should be considered, such as the LOS probability between the two terminals, the reflection coefficient of the RIS component, the far and near field conditions, and the operating frequency.

C. RIS Features Suitable for Positioning
The RIS has the following features which make it has a good engineering application prospect.
Reconfigurable and Intelligent (Customize Wireless Environment): The significant advantage of the surface composed of dynamic meta-atoms is that it can be reconfigured after manufacturing and deployment in the environment. RIS configuration is performed by programming, so that each element continuously reconfigures the amplitude, phase, frequency, and polarization of the incident signal in real time, and has the ability to apply specific transformations to impact radio waves, thereby effectively controlling the propagation environment.
Focusing Energy: Utilizing the reconfigurable characteristics of RIS, electromagnetic waves can be effectively designed to steer them in any target user direction. The ability to focus energy in the 3-D space makes RIS an attractive technology for positioning applications [41].
Enlarge Coverage Area: Above 6 GHz, severe signal attenuation and blocking inhibit the long-distance transmission of millimeter waves, and the positioning performance of celledge users is degraded. RIS bypasses obstacles by intelligently reflecting signals to create an adaptable virtual line of sight connection in blind areas or low coverage areas where LOS communication is impossible or insufficient LOS communication. The coverage extension in mm-wave communications affected by indoor congestion is particularly useful [42].
Low Cost (Low Manufacturing Cost): In the actual implementation of RIS, its components are usually cheap, such as reflective components, miniature antennas, and diodes.
Low Deployment Cost: Due to the small size and area of the RIS, it is easy to deploy (or remove) in buildings, ceilings, walls, and exterior walls, with low deployment and maintenance costs. In addition, in the integration of RIS into the existing communication network, only the network protocol needs to be changed, without the need to adjust the hardware facilities, so that RIS can be deployed and integrated into the wireless network at a lower cost.
Nearly Passive: The RIS does not require any active elements and complicated computation, such as active sensors, baseband processing units, and RF chain [43]. The RIS itself does not need a power supply to complete the signal reflection and only requires extremely low power when the RIS is configured. Liaskos et al. [30] proposed an approach to make the wireless environments programmable with the metasurface. This method requires very little energy consumption. For example, if a RIS with the meta-atoms size of 8 mm × 8 mm is used to cover a wall of 3 m × 5 m, even if all diodes are in the "ON" state, the energy consumption required for the entire wall is only 125 mW/m 2 at the maximum.
III. RIS-ASSISTED 6G IOT WIRELESS POSITIONING RIS can provide high-precision, wide-coverage location services for IoT systems. In order to facilitate readers to read and understand the correlations and differences between documents, and to facilitate access to related documents, Table II is provided.

A. RIS for IoT Communications
Before demonstrating the capabilities of RIS-assisted IoT positioning, we first review the application of RIS in IoT communication to motivate the deployment of RIS in IoT. 6G envisages a four-layer network architecture, namely, a space-air-ground-underwater/sea network, to provide ubiquitous connections and support new vertical areas of the IoT in the future. The space network tier consists of various satellites. The provision of IoT services over satellite networks has been investigated by the research community [44], [45]. Tekbıyık et al. [46] introduced an RIS-assisted LEO satellite framework for energy-efficient IoT. It is recommended to use RIS on satellites and use RIS units to mitigate path losses related to long-distance transmission. Research shows that with the assistance of RIS, satellites can provide up to 10 5 times higher downlink and achievable uplink rate for IoT networks. The air network tier consists of various flying BSs, such as high-altitude platforms, mobile airborne units, UAVs, etc. In recent years, UAVs have been widely used in various IoT scenarios [47]. Ranjha and Kaddoum [48] relied on UAV and RIS to transmit short URLLC command packets between ground IoT devices. It is pointed out through simulation that if the position of the UAV is placed reasonably, ultrahigh reliability can be achieved by increasing the number of RIS antenna elements, which proves that RIS performance gains. For 6G IoT, ensuring privacy and security is the key challenge in the future. Vo et al. [49] studied the physical-layer security problem in the UAV-RIS-assisted IoT cognitive radio network with eavesdroppers. The numerical results show that with the increase in the number of RIS units, the security performance of the network is improved. For the general ground network tier, Makarfi et al. [50] studied IoT networks using RIS under complex fading and shadowed generalized channels, with source nodes employing RIS-based access points. The obvious benefits of the RIS-assisted IoT in  [51] introduced RIS to solve the communication and power efficiency problems of short-range and low-power IoT applications. Compared with the traditional ambient backscatter communication (AmBC) method, it has superior performance in terms of error rate and data rate.
Mursia et al. [52] proposed a method to jointly optimize the precoding strategy of the BS and the RIS parameters to help access the large-scale IoT based on millimeter wave technology. Simulation results show that this method can improve the performance gain of the sum rate. Liu et al. [53] introduced RIS to assist high-speed rail communication to support highspeed rail IoT services. The authors propose three potential deployment schemes of RIS and analyze the actual challenges and future research directions. The underwater network tier includes submarines, autonomous underwater vehicles (AUVs), ships, etc., which have been proposed to establish underwater IoT to provide sufficient connectivity for wireless networks in underwater media. RIS use cases in challenging environments such as underwater are considered in [54]. Three placement methods of RIS are given: 1) attached to the ground or shore; 2) attached to the automatic underwater vehicle; and 3) floating. RIS is used to combat the extreme frequency selectivity caused by the multipath propagation of underwater acoustic signals, thereby increasing the effective signal bandwidth. In addition, Kisseleff et al. [54] discussed potential use cases and deployment strategies of RIS in supporting underground IoT, industry 4.0, and emergency networks, and paved the way for its system design. Given the potential of RIS to reduce frequency selectivity, and improve coverage, reliability, and energy efficiency of communication networks, RIS is expected to become a key driver of future CEs wireless systems.

B. Potential of RIS-Assisted Positioning
Next-generation cellular networks can witness the creation of SREs, where walls and objects will cover RISs to enhance communication and positioning performance. RIS can interact with the environment to effectively reconstruct radio signals and change their natural properties, including direction, polarization, and more, [30], thereby converting the wireless channel into an intelligent transmission entity [55]. The ability to intelligently control the propagation channel of wireless signals makes it possible for energy to be concentrated in a 3-D space for transmission and reception, which will bring brandnew capabilities for communication, sensing, and control of the electromagnetic environment [56]. Wymeersch et al. [18] pointed out that as long as appropriate models and algorithms can be developed, RISs can improve positioning accuracy and help expand physical coverage, thereby facilitating positioning. Here, we summarize the potential and source of RIS positioning.
First, the much larger surface area of RIS gives it a significant advantage over traditional MIMO positioning in terms of positioning. On the one hand, the RIS uses the entire continuous surface to transmit and receive radio signals. On the other hand, as the surface area of the RIS increases, the Cramer-Rao lower bound (CRLB) for terminal positioning shows a decreasing trend [41], [57]. The author separately derives the location CRLB of the terminal on the central perpendicular line (CPL) of the RIS surface or away from the CPL, focusing on a direct view of the CRLB in relation to the surface area of the RIS. Theoretical analysis and simulation results both corroborate the following conclusions, except that when the terminal is on the CPL, the coordinate in the dimension of the vertical RIS plane decreases linearly with the surface area of the RIS. In other cases, the CRLB generally decreases with the surface area of the RIS quadratically. The author also certifies that the distributed deployments, that is, split the RIS into more small pieces, are competent to enlarge the coverage of localization, so that can improve the average CRLB.
Second, quantifying the phase and amplitude can keep the RIS positioning at reasonable costs. The neighboring previous article is under the assumption of full-resolution phase and amplitude, [58] is an extension of it, and studies the effect of different phase and amplitude quantization resolution combinations on CRLB. The numerical results of the simulation show that the CRLB loss caused by the amplitude quantization can be ignored and is close to the full resolution results, while the influence of the phase quantization is far greater than the amplitude quantization. This motivates the actual implementation of RIS-assisted positioning, which can focus on improving the phase resolution to achieve better positioning accuracy. Hu et al. [41], [57], and Alegrla and Rusek [58] researched the potential of RIS terminal positioning under the assumption of perfect LOS propagation, and theoretically analyze the CRLB of positioning estimation. Although the NLOS situation is more in line with the realistic scenario, the analysis of the CRLB obtained in the LOS situation is of reference significance for understanding the positioning performance of the RIS-assisted terminal in the NLOS situation.
Third, Huang et al. [59] proposed an efficient online wireless RIS optimal phase configuration scheme based on a deep learning method, which enables the transmission energy to be focused on the expected user position with high accuracy and maximizes the user's received signal strength. The high focusing capability of the large geometric size RIS can be used to finely estimate the position of mobile terminals and devices, thereby supporting high-precision ranging and radio positioning. Wymeersch and Denis [60] analyzed an RIS-aided downlink positioning problem from the Fisher Information perspective and shows coverage and accuracy gains in comparison with reflecting surface and scatter point.

C. RIS-Assisted mm-Wave MIMO Positioning Systems
Because the large bandwidth in the mm-wave band results in high temporal resolution and the equipped large antenna arrays with extremely narrow beams provide high spatial resolution in the angular domain, positioning with signals in the mm-wave and MIMO systems are widely studied [61], [62], [63], [64]. Research shows that a single BS can also obtain promising positioning accuracy [65]. The following is a summary of the results of RIS-assisted mm-wave MIMO systems.
In [65], RIS is introduced into the mm-wave MIMO positioning system as a reflector, assuming that the LOS path between the BS and the MS exists. By deriving the Fisher information matrix (FIM), the CRLB of the standard deviation of the positioning estimation error and the orientation estimation error are obtained, which demonstrates that the RISassisted mm-wave MIMO positioning system is superior to the traditional positioning system. However, the positioning under the condition of obscured LOS has not been greatly discovered. The following year, the same author further studied the adaptive beamforming of RIS-assisted mm-wave MIMO positioning in a similar positioning scenario as above [43]. The difference is that the direct loss between the BS and the MS is supposed blockage. The author proposes an adaptive phase shifter design based on a hierarchical codebook (HCB) and receiver feedback to optimize the phase value of the discrete RIS unit and improve the performance in terms of positioning accuracy and data rate. It is worth mentioning that this research has reference significance for the study of joint positioning and communication. Liu et al. [66] derived the FIM and CRLB to estimate the absolute position of the MS in the scene where the LOS is blocked. By optimizing the design of the reflect beamforming to minimize CRLB, the positioning accuracy can achieve the decimeter level or even the centimeter level. In view of the LOS blockage,Čišija et al. [67] studied a multiuser location algorithm based on an HCB. The simulation results under different SNR conditions show that the proposed method has the potential to assist multiuser positioning in mm-wave MIMO radar systems under a suitable HCB design. Aiming at the signal blocking problem in mmWave MIMO communication scenarios, Wang and Zhang [68] proposed a method of joint beam training and positioning. The channel parameter Angle of Departure (AoD) is estimated by beam training and used by the mobile terminal to estimate its position, and the position information in turn is beneficial to the refinement of beam training parameters. Simulation results show that the proposed scheme can satisfy multiuser centimeter-level positioning accuracy. Zhang et al. [69] proposed a two-stage positioning method with dual RISs. In the first stage, the phase shift of the reflective element is designed for each RIS, and the position information is estimated in the second stage. Numerical results show that the positioning accuracy of this method can reach 10 −5 -10 −4 m, which is an order of magnitude lower than that of a single RIS auxiliary method.

D. RIS-Assisted Indoor Localization
Nowadays, GPS can serve outdoor positioning with acceptable accuracy. However, in an environment where GPS is rejected, such as indoors, coupled with the adverse effects of abundant obstacles on the signal propagation, the positioning becomes unfavorable. The introduction of RIS in the GPS rejection environment can not only overcome the communication congestion caused by obstacles but also assist the high-precision positioning of users in the environment.

1) Assist RSS-Based Positioning:
With the ability of simple implementation, the RSS positioning technology using widely. RIS can be employed to: 1) enhance the strength of the received signal by reusing valuable signals in the multipath environment and 2) reduce and avoid co-channel interference by selecting different propagation paths. Accordingly, positioning algorithms based on the received signal strength can greatly benefit from RIS to enhance the received signal strength and reduce co-channel interference. However, its accuracy is limited due to the difficulty of distinguishing  adjacent RSS values. Zhang et al. [75] through theoretical analysis and actual measurement, proved that RIS can indeed configure the wireless environment. As shown in Fig. 7, the different configurations of each RIS element are distinguished by digital marks. The colored block in the figure represents the RSS measurement value at the corresponding position. Comparing the two pictures, it is obvious that the RSS measurement values at the same position change with the RIS configuration. The ability to manually manipulate the wireless environment provides an opportunity to solve the problem of indistinguishable RSS values from adjacent locations in RSS indoor positioning. Fig. 8 presents a system model for RIS-assisted indoor positioning. In order to select the most suitable RIS configuration to provide high-precision positioning results, a new formula for the RIS-assisted positioning problem is presented [14], which minimizes the weighted probability of incorrect positioning, and an iterative configuration optimization algorithm is designed to solve the customized problem. Numerical results show that the positioning error of the proposed method is much smaller than that of the positioning method without RIS assistance. In the same year, the author also proposed a phase shift optimization Fig. 9. Fraunhofer distance of RIS with size D under incident waves at (a) sub-6 GHz and (b) millimeter-wave frequencies [19]. algorithm (PSO) to solve the same problem [70]. Compared with the traditional RSS-based solution, the positioning error of this solution can be reduced by at least three times, providing a new solution to the multiuser positioning problem based on RSS. Nguyen et al. [71] proposed a fingerprint location method based on RSS value, which uses machine learning to reduce the complexity of generating environmental maps, thereby improving the accuracy and efficiency of positioning.
2) Assist UWB-Based Positioning: If the RIS reflection phases satisfy θ 1 = θ 2 = · · · = θ n ∈ [0, 2π), the multipath signal passing through each RIS unit can be labeled, which provides a feasible idea for processing multipath signals. Ma et al. [27] developed a novel indoor RIS-assisted positioning scheme, combining the ability of RIS to tag multipath channels and the high multipath resolution of UWB signals. The CRLB of the proposed positioning scheme is derived, which exhibits that RIS has the potential to realize accurate positioning with a single access point. In addition, because the proposed scheme only requires a single access point and some low-cost RIS units, it provides a more accurate and cost-effective solution for indoor positioning.

E. Near-Field Localization
According to the definition, the signal beyond the Fraunhofer distance (R = 2D 2 /λ) is located in the far field, which satisfies the plane wave assumption [18], where D is the RIS size, and λ is the incident wavelength, and the nearfield channel is under the spherical wave model. It is observed that the range of the near-field area of RIS is inversely proportional to the incident wavelength. Fig. 9 shows the variation of the Fraunhofer distance to the frequency of the incident wave under different RIS sizes. For a 1-m 2 RIS, when working at 30 GHz, the near field area is about 200 m [76]. In particular, as the operating frequency increases and the surface area increases, the RIS radiation near-field range expands, and the UE is likely to be located in the near field. Therefore, the farfield assumption is not always correct for the RIS auxiliary system, which makes the positioning inapplicable in certain scenarios, such as indoor environments.
The wavefront curvature of the transmitted signal can be used for localization under the conditions of limited infrastructure and hardware complexity. The possibility of using the mm-wave wavefront curvature to locate the transmitter has been researched in [77]. Abu-Shaban et al. [72] proposed a two-stage positioning algorithm to realize the positioning of the transmitter when RIS is used as a lens. For a medium-sized RIS operating at 28 GHz, decimeter-level positioning can be achieved within 3 m of the near field. Elzanaty et al. [16] established a general model of near-field far-field positioning and proposes a phase design based on SNR to minimize CLRB. Compared with the traditional system without RIS, the proposed scheme based on the SNR can achieve two orders of magnitude and one order of magnitude reduction in positioning error bounds and directional error bounds, respectively. The aforementioned RIS near-field positioning works ignore the realization of blockage, but it is necessary to discuss to evaluate the performance of positioning. Rahal et al. [73] studied the near-field positioning potential of the single-input single-output (SISO) system under a severe line of sight obstruction. Dardari et al. [74] proposed a two-step positioning algorithm based on TOA, and the simulation verified the possibility of maintaining high positioning accuracy even in the case of severe obstruction in the near-field area of RIS. Considering the knowledge of the RIS amplitude model, Ozturk et al. [33] proposed a low complexity near-field localization algorithm called approximate MML (AMML) by using the Jacobi-Anger expansion. Using the result as the initial position estimate, and further proposes an iterative refinement algorithm for joint localization and RIS amplitude model parameters update. The simulation results show the effectiveness of the proposed low-complexity localization algorithm, and the localization accuracy of the iterative algorithm is asymptotically reaching CRLB.

IV. CHALLENGES AND ROAD AHEAD
Above we reviewed the existing research on RIS positioning. Introducing RIS into the IoT can improve positioning accuracy with the advantages of low cost and low energy consumption, which is a promising trend. Here, we emphasize the specific challenges of RIS-assisted 6G IoT positioning and valuable potential research directions.
A. Challenge of 6G-IoT Positioning System 1) Large-Scale IoT Devices Positioning: From the above research on RIS-assisted positioning, it can be found that even if multitarget positioning is considered, it can only be effective for a few targets. Due to the high growth rate of the number of IoT devices each year, it can be predicted that the future 6G IoT devices support will be far more than the current 5G, and the devices with positioning requirements in various indoor and outdoor IoT applications, such as smart homes and smart cities will become dense. Further research on the ability of RIS to assist multiuser localization in the IoT is more meaningful for the practical application of RIS. For large-scale user localization, it is necessary to design a reasonable multiuser localization protocol, RIS beamforming, and localization algorithm. Even network architectures need to be redesigned to support ultramassive machine-type communications.
2) Challenges of Higher Frequency Bands for 6G: In moving from 5G to 6G, the signal frequency used will increase from mmWave band to THz even higher. On the one hand, high-frequency transmission is more sensitive to the obstruction of obstacles in the propagation environment, which makes the impact of NLOS on positioning more serious. In the indoor positioning scene, most of the current studies are conducted under the assumption of LOS diameter, so it is necessary to design the system model in the real environment for the future 6G IoT indoor application. On the other hand, the increase in frequency leads to changes in the size of the RIS component. Therefore, component manufacturing and integration become challenging. The challenge of how to adapt to higher frequency bands is an issue that needs to be discussed, such as whether it is necessary to design new hardware structures and working principles.
3) Inevitable Near-Field Positioning: As shown in Fig. 9, with the increase of carrier frequency in 6G, the range of the near-field region will be expanded by orders of magnitude, and near-field positioning is inevitable in the future 6G network. In the near-field condition, the phase of the spherical wave is a nonlinear function of the incident angle and distance information of each path. Considering the distance between BS-RIS and RIS-UE in the RIS-assisted IoT system, the equiphase surfaces become elliptical [78]. Due to the different electromagnetic characteristics of spherical waves and plane waves, the existing technology based on far-field assumptions will suffer huge performance loss when applied directly to near field. Introducing RIS into the IoT positioning system requires reconsidering the channel model on the basis of the elliptical phase surfaces, and providing accurate beam focusing for users at different angles and distances.
B. Road Ahead 1) Deploying Distributed RIS: It is mentioned in [57] that dividing the entire large RIS into smaller pieces and deploying them distributedly at a certain distance can obtain a more robust overall localization performance than centralized deployment and improve the average localization performance. Kisseleff et al. [54] believed that RIS distributed operation is very useful for disaster environments. Multiple RIS patches that can be easily separated from the complete RIS can improve connectivity and deal with the fragmentation of infrastructure caused by disasters. Distributed deployment is at the cost of increasing complexity, requiring that each separable RIS patch is intelligent enough and the feedback overhead is large. In addition, RIS design should be especially carried out for detachable requirements.
2) Integrated Positioning and Communication: Presently, a large number of researches related to RIS focus on the RIS auxiliary communication network, and studies have shown that the introduction of RIS into the IoT can improve communication performance. The potential of RIS-assisted positioning has also been discovered in recent years, which makes it very meaningful for RIS to achieve integrated positioning and communication in the IoT. By name, integrating, location and communication mean that the entire process shares the same set of hardware architectures and algorithms, which minimizes the deployment cost of the location network. However, how to jointly design RIS to serve communication and positioning has not been widely studied. Existing research on RIS-assisted integrated positioning and communication is limited, and the most valuable is [43], as mentioned above. In addition, the closed RIS phase profile proposed in [16] is very suitable for joint positioning and communication. When positioning is involved, the transmission of signaling and data packets for positioning will interfere with communication and reduce the transmission efficiency of communication data. This needs to be better considered according to the goal of practical application. The theoretical limits of basic information, network architecture and transmission protocols, the signal processing technology at high frequency and the coding/modulation/beamforming design of RIS-assisted integrated positioning and communication are the directions for further research in the future.
3) Near-Field Beam Focusing: The near-field propagation of electromagnetic wave signals in a 6G system is both a challenge and an opportunity. The far-field plane wave can only steer the beam energy to a specific angle. Due to the interference between users at a similar angle, the ability to serve dense users in the same angle domain under far-field conditions is limited. However, the near-field wavefront has an extra resolution in the range domain, so it has the ability to focus the beam energy at a specific angle and distance, which provides the possibility for more accurate positioning of multiusers.

4) User Positioning in Tunnels:
Intelligent transportation systems (ITSs) are an important part of future smart cities, and autonomous vehicles (AVs) equipped with various types of sensors and network access units will play an important role in future intelligent transportation vehicles [79]. The GPS signal strength in the highway tunnel environment is weak, and the positioning in the tunnel has always been challenging. For future AVs, accurate positioning is more secure. In addition, the accurate positioning of personnel or vehicles in the event of tunnel accidents is of great significance for efficient rescue work. Chen and Pan [80] introduced RIS into the tunnel environment and verifies the effectiveness of RIS in reducing the blocking probability. Deploying RIS on the wall or ceiling of a tunnel, RIS has the ability to increase received signal strength by reusing valuable signals in a multipath environment, potentially enhancing positioning performance. The challenge of practical application is the high-speed mobility of the vehicle, which requires that the RIS controller has sufficient fast computing ability to adjust the direction of the reflected beam to support the real-time positioning of the vehicle, which has not been well considered in the existing RIS positioning literature. When locating a vehicle moving in a tunnel, the RIS needs to be able to estimate the channel in real time as the target user moves, and design an appropriate phase shifter. When perfect channel state information is not available, it can be further investigated whether only channel state statistics are available.

V. CONCLUSION
The construction of the 6G wireless communication system puts forward new requirements for positioning and brings new challenges. Using RIS to improve the radio location of IoT devices is a promising research direction. This article first introduces the working principle of RIS and summarizes the two major advantages of RIS for assisting 6G IoT wireless positioning, namely, reconfiguration and low cost. Then, we focus on the classification and review of the existing RIS-assisted positioning techniques, which prove the potential of RIS-assisted positioning. In different scenarios with the assistance of RIS, the positioning performance of mmwave MIMO positioning, indoor positioning, and near-field positioning can be improved. Finally, this article discusses the challenges of RIS-assisted future 6G IoT positioning in large-scale user positioning, using higher frequency bands and near-field positioning. Encouragingly, using distributed RIS deployment, integrating RIS-assisted positioning and communication, studying beam focusing in the near field, and using RIS-assisted AV positioning are the future development directions. It is hoped that the summary of the current RIS-assisted positioning progress in this article can provide valuable guidance for researchers engaged in RIS positioning or 6G wireless network positioning.