ABSTRACT Reducing the running time of graph algorithms is vital for tackling real‐world problems such as shortest paths and matching in large‐scale graphs, where path information plays a crucial role. To address this critical challenge, this paper introduces a graph restructuring algorithm that identifies bipartite cliques and replaces them with tripartite graphs. This restructuring leads to fewer edges while preserving complete graph path information, enabling the direct application of algorithms like matching and all‐pairs shortest paths to achieve significant runtime reductions, especially for large, dense graphs. The running time of the proposed algorithm for a graph , with and is , which is better than , the running time of the best existing algorithm for speeding‐up other graph algorithms (the Feder–Motwani ( FM ) algorithm), where . Both the FM algorithm and the proposed algorithm are originally formulated for bipartite graphs, but can also be applied to general directed or undirected graphs. Our extensive experimental analysis demonstrates that the proposed algorithm achieves up to 21.26% higher reduction in the number of edges and runs up to faster than the FM algorithm. On large synthetic graphs with up to 1.05 billion edges, it attains a reduction in the number of edges of up to 74.36%. On real‐world graphs, it achieves a reduction in the number of edges by up to 46.8%. Furthermore, when used as a preprocessing step, our approach yields up to a speedup for the matching algorithms on large synthetic graphs, and up to a speedup for the All‐Pairs Shortest Path algorithms on real‐world graphs, when compared to using the given graph as input.

Recent advances in mobile System-on-Chip (SoC) architectures–equipped with specialized AI accelerators such as GPUs, DSPs, and NPUs–have made on-device inference a viable alternative to edge-assisted processing. Mobile Augmented Reality (MAR) applications increasingly rely on multiple deep learning models to enable real-time interaction with virtual objects. Although several works aim to enhance MAR performance, they often overlook the runtime contention between AI inference and AR rendering tasks, which degrades both object quality and AI latency. In this paper, we introduce HBO+, a runtime framework for MAR applications that jointly controls virtual object triangle count and AI task placement on local and edge resources. By combining Bayesian optimization with heuristic algorithms, HBO+dynamically balances AI task performance and virtual object quality to improve overall system performance. We implement HBO+on Android and evaluate it using commercial smartphones and edge servers. The experimental results show that our method reduces the average latency of the AI task by up to 1.38× and improves the average virtual object quality by up to 30% compared to state-of-the-art baselines.

Ensuring low battery degradation in Autonomous Ground Robot (AGR) fleets operating in online environments (e.g., delivery services) is essential for enhancing their long-term sustainability. However, most existing studies either rely on offline methods—unsuitable for scenarios requiring real-time decisions—or focus solely on maximizing task allocation, resource utilization, or revenue, with limited consideration for battery health. Additionally, maximizing fleet sustainability requires bounded relative revenue losses from unassigned tasks within a user-defined acceptable limit to make it an attractive option for industry. To address these limitations, we propose an online task and charge allocation framework that jointly optimizes revenue generation and battery lifespan, while allowing users to explicitly constrain relative revenue losses. The framework includes three event-driven algorithms: BTC-M, which computes optimal decisions at each event, and two computationally efficient greedy variants, BTC-G and BTC-WG, which provide sub-optimal solutions with reduced overhead. We evaluate the performance of our approach under different task arrival distributions representative of real-world applications. Simulation results based on a real AGR, compared against multiple baselines, demonstrate that our framework can extend battery lifespan by up to 19% with minimal revenue loss.

In geographical image segmentation, performance is often constrained by the limited availability of training data and a lack of generalizability, particularly for segmenting mobility infrastructure such as roads, sidewalks, and crosswalks. Vision foundation models like the Segment Anything Model (SAM), pre-trained on millions of natural images, have demonstrated impressive zero-shot segmentation performance, providing a potential solution. However, SAM struggles with geographical images, such as aerial and satellite imagery, due to its training being confined to natural images and the narrow features and textures of these objects blending into their surroundings. To address these challenges, we propose Geographical SAM (GeoSAM), a SAM-based framework that fine-tunes SAM using automatically generated multi-modal prompts. Specifically, GeoSAM integrates point prompts from a pre-trained task-specific model as primary visual guidance, and text prompts generated by a large language model as secondary semantic guidance, enabling the model to better capture both spatial structure and contextual meaning. GeoSAM outperforms existing approaches for mobility infrastructure segmentation in both familiar and completely unseen regions by at least 5% in mIoU, representing a significant leap in leveraging foundation models to segment mobility infrastructure, including both road and pedestrian infrastructure in geographical images. The source code is publicly available.

Mobile Augmented Reality (MAR) must balance AI latency and visual quality of virtual objects amid user and scene dynamics. We present an edge-assisted RL framework that jointly optimizes AI task allocation across CPU, GPU, and NPU and the triangle ratio of rendered virtual objects. A lightweight on-device context predictor forecasts near-future states, while a multi-actor, single-reward model on the edge makes decisions via a one-step contextual bandit with UCB-based actor selection. This design enables proactive, context-aware adaptation and fast policy reuse. Experimental tests on real smartphones and edge servers show that our approach reduces exploration by up to 57.1%, extends exploitation up to 12×, and improves reward by 193% over state-of-the-art baselines.

The deployment of deep neural networks (DNNs) on edge devices is becoming increasingly common in latency-sensitive applications such as autonomous driving, real-time video analytics, and augmented reality. However, modern DNNs are rapidly growing in complexity, and edge GPUs often lack the computational resources available in cloud counterparts. This leads to increased inference latency and challenges in meeting strict Service Level Agreements (SLAs).

When mobile apps are used extensively in our daily lives, their responsiveness has become an important factor that can negatively impact the user experience. The long response time of a mobile app can be caused by a variety of reasons, including soft hang bugs or prolonged user interface APIs (UIAPIs). While hang bugs have been researched extensively before, our investigation on UI-APIs in today's mobile OS finds that the recursive construction of UI view hierarchy often can be time-consuming, due to the complexity of today's UI views. To accelerate UI processing, such complex views can be preprocessed and cached before the user even visits them. However, pre-caching every view in a mobile app is infeasible due to the incurred overheads on time, energy, and cache space. In this paper, we propose MAPP, a framework for Mobile App Predictive Pre-caching. MAPP has two main modules, 1) UI view prediction based on deep learning and 2) UI-API pre-caching, which coordinate to improve the responsiveness of mobile apps. MAPP adopts a per-user and per-app prediction model that is tailored based on the analysis of collected user traces, such as location, time, or the sequence of previously visited views. A dynamic feature ranking and model selection algorithm is designed to judiciously filter out less relevant features for improving the prediction accuracy with less computation overhead. MAPP is evaluated with 61 real-world traces from 18 volunteers over 30 days to show that it can shorten the response time of mobile apps by 59.84 % on average with an average cache hit rate of 92.55 %.

Edge computing has been developed as a low-latency data driven computation paradigm close to the end user to maximize profit, and/or minimize energy consumption. Edge computing allows each user’s task to analyze locally-acquired sensor data at the edge to reduce the resource congestion and improve the efficiency of data processing. To reduce application latency and data transferred to edge servers it is essential to consider data sharing for some user tasks that operate on the same data items. In this article, we formulate the data sharing-aware allocation problem which has as objectives the maximization of profit and minimization of network traffic by considering data-sharing characteristics of tasks on servers. Because the problem is${\sf NP-hard}$, we design the${\sf DSTA}$algorithm to find a feasible solution in polynomial time. We investigate the approximation guarantees of${\sf DSTA}$by determining the approximation ratios with respect to the total profit and the amount of total data traffic in the edge network. We also design a variant of${\sf DSTA}$, called${\sf DSTAR}$that uses a smart rearrangement of tasks to allocate some of the unallocated tasks for increased total profit. We perform extensive experiments to investigate the performance of${\sf DSTA}$and${\sf DSTAR}$, and compare them with a representative greedy baseline that only maximizes profit. Our experimental analysis shows that, compared to the baseline,${\sf DSTA}$reduces the total data traffic in the edge network by up to 20% across 45 case study instances at a small profit loss. In addition,${\sf DSTAR}$increases the total profit by up to 27% and the number of allocated tasks by 25% compared to${\sf DSTA}$, all while limiting the increase of total data traffic in the network.

The emergence of modern mobile System on Chips (SoCs), featuring robust neural network accelerators such as GPUs, DSPs, and NPUs, has made on-device inference a compelling alternative to cloud-assisted inference. Typical mobile augmented reality (MAR) applications enable users to interact with virtual objects, leveraging diverse Artificial Intelligence (AI) capabilities facilitated by a range of deep learning models. Several studies seek to improve the performance of MAR apps. However, they often overlook the computational concurrency between the AR tasks necessary to render virtual objects and the AI tasks, which can influence virtual object quality and AI inference response time. In this paper, we present HBO, a framework for MAR apps that trades off between AR and AI task performance. HBO leverages Bayesian optimization and heuristic algorithms to jointly manipulate the virtual objects' triangle count and AI task allocation for optimized MAR app performance. We have implemented HBO on Android and tested it on real smartphones and with real users. Our results show that HBO helps reduce the average AI task latency by up to 3.5x and increase the average virtual object quality by up to 38.7% compared to several stateof-the-art baselines.

Energy harvesting sources, such as solar, wind, or vibration, combined with rechargeable batteries, are a promising way to power Low-Power Wide-Area Network (LPWAN) devices to reduce the cost and frequency of redeploying single-use batteries. However, being oblivious to the usage of rechargeable batteries can severely reduce their capacity to store energy, which is also known as the battery lifespan. Existing energy-aware protocols mostly focus on network lifetime and pay little attention to maximizing the battery lifespan of network nodes while the latter can directly help reduce battery waste and enhance environmental sustainability. In this paper, we propose the first Media Access Control (MAC) protocol to maximize the minimum battery lifespan among all nodes in an LPWAN based on LoRa. Our approach differs from traditional objectives focusing on min-imizing energy consumption or maximizing network lifetime as they may not necessarily maximize battery lifespan. The proposed MAC protocol leverages the concept of software-defined batteries to regulate the energy stored and consumed by each node's battery based on estimated energy requirements, green energy generation, and changes in data utility. To limit the degradation of battery capacity due to continuous charging/discharging, the underpinning idea is to determine an appropriate time for each transmission considering its impact on battery degradation while also minimizing the impact on data utility. Furthermore, the energy stored in each battery is limited to reduce calendar aging, the natural degradation of battery capacity over time. The proposed protocol is local, online, and asynchronous, and incurs low overhead. We evaluate our approach through experiments on a LoRa network and large-scale simulations in NS-3. The experiments show that the proposed MAC protocol improves battery lifespan by up to 69.7% and data utility by up to 39% in a current LoRa network while incurring a CPU utilization overhead of only 12 % at each LoRa node.

Autonomous mobile robots (AMRs) are capable of carrying out operations continuously for 24/7, which enables them to optimize tasks, increase throughput, and meet demanding operational requirements. To ensure seamless and uninterrupted operations, an effective coordination of task allocation and charging schedules is crucial while considering the preservation of battery sustainability. Moreover, regular preventive maintenance plays an important role in enhancing the robustness of AMRs against hardware failures and abnormalities during task execution. However, existing works do not consider the influence of properly scheduling AMR maintenance on both task downtime and battery lifespan. In this paper, we proposeMTC,a maintenance-aware task and chargingscheduler designed for fleets of AMR operating continuously in highly automated environments.MTCleverages Linear Programming (LP) to first help decide the best time to schedule maintenance for a given set of AMRs. Subsequently, the Kuhn-Munkres algorithm, a variant of the Hungarian algorithm, is used to finalize task assignments and carry out the charge scheduling to minimize the combined cost of task downtime and battery degradation. Experimental results demonstrate the effectiveness ofMTC, reducing the combined total cost up to 3.45 times and providing up to 68% improvement in battery capacity degradation compared to the baselines.

Private and public transportation will be dominated by Autonomous Vehicles (AV), which are safer than regular vehicles. However, ensuring good performance for the autonomous features requires fast processing of heavy tasks. Providing each AV with powerful computing resources may result in increased AV cost and decreased driving range. An alternative solution is to install low-power computing hardware on each AV and offload the heavy tasks to powerful nearby edge servers. In this case, the AV’s reaction time depends on how quickly the navigation tasks are completed in the edge server. To reduce task completion latency, the edge servers must be equipped with enough network and computing resources to handle the vehicle demands, which show large spatio-temporal variations. Thus, deploying the same resources in different locations may lead to unnecessary resource over-provisioning. In this paper, we leverage simulations using real traffic data to discuss the implications of deploying heterogeneous resources in different city areas to sustain peak versus average demand of edge-assisted AVs. Our analysis indicates that a reduction in network bandwidth and computing cores of up to 60% and 50%, respectively, is achieved by deploying edge resources for the average demand rather than peak demand. We also investigate how the peak-hour demand affects the safe travel time of AVs and find that it can be reduced by approximately 20% if they would be rerouted to areas with a lower edge-resource load. Thus, future research must consider that traditional turn-by-turn navigation systems may not provide the fastest routes for edge-assisted AVs.

In Vehicular Edge Computing (VEC) systems, the computing resources of connected Electric Vehicles (EV) are used to fulfill the low-latency computation requirements of vehicles. However, local execution of heavy workloads may drain a considerable amount of energy in EVs. One promising way to improve the energy efficiency is to share and coordinate computing resources among connected EVs. However, the uncertainties in the future location of vehicles make it hard to decide which vehicles participate in resource sharing and how long they share their resources so that all participants benefit from resource sharing. In this paper, we propose VECMAN, a framework for energy-aware resource management in VEC systems composed of two algorithms: (i) a resource selector algorithm that determines the participating vehicles and the duration of resource sharing period; and (ii) an energy manager algorithm that manages computing resources of the participating vehicles with the aim of minimizing the computational energy consumption. We evaluate the proposed algorithms and show that they considerably reduce the vehicles’ computational energy consumption compared to the state-of-the-art baselines. Specifically, our algorithms achieve between 7 and 18 percent energy savings compared to a baseline that executes workload locally and an average of 13 percent energy savings compared to a baseline that offloads vehicles’ workloads to RSUs.

Mobile Augmented Reality (MAR) apps may cause short battery life due to high-quality virtual objects rendered in the augmented environment. State-of-the-art solutions propose to balance energy consumption and user-experience using a static set of decimated object versions within the app. However, they do not consider that each object has unique characteristics, which highly influence how the user-perceived quality changes according to user-object distance and triangle count. As a result, they may lead to limited energy savings, a high storage overhead, and a high burden on the MAR app developer. In this paper, we propose eAR, an edge-assisted autonomous and energy-efficient framework for MAR apps designed to solve the limitations of state-of-the-art solutions. eAR features an offline software running on an edge server that leverages Image Quality Assessment (IQA) to model user-perceived quality for each virtual object in terms of triangle count and user-object distance. In addition, eAR features a runtime lightweight optimization algorithm that dynamically decides the most energy-efficient virtual object triangle count to request from the edge server based on (i) the per-object models of user-perceived quality, (ii) energy consumption models for mobile GPU and network interface, and (iii) a user path prediction system that estimates near-future user-object distances. eAR is completely autonomous and can be easily integrated into most MAR apps as an open-source library. Our results show that eAR can help reduce energy consumption by up to 16.5% while reducing storage overhead by almost 60% compared to existing schemes, with minimal MAR app developer effort and minimal impact on user-perceived quality.

Bias in the training data can jeopardize fairness and explainability of deep neural network prediction on test data. We propose a novel bias-tailored data augmentation approach, Counterfactual Interpolation Augmentation (CIA), attempting to debias the training data by d-separating the spurious correlation between the target variable and the sensitive attribute. CIA generates counterfactual interpolations along a path simulating the distribution transitions between the input and its counterfactual example. CIA as a pre-processing approach enjoys two advantages: First, it couples with either plain training or debiasing training to markedly increase fairness over the sensitive attribute. Second, it enhances the explainability of deep neural networks by generating attribution maps via integrating counterfactual gradients. We demonstrate the superior performance of the CIA-trained deep neural network models using qualitative and quantitative experimental results. Our code is available at: https://github.com/qiangyao1988/CIA

Mobile artificial intelligence has recently gained more attention due to the increasing computing power of mobile devices and applications in computer vision, natural language processing, and internet of things. Although large pre-trained language models (e.g., BERT, GPT) have recently achieved the state-of-the-art results on text classification tasks, they are not well suited for latency critical applications on mobile devices. Therefore, it is essential to design tiny models to reduce their memory and computing requirements. Model compression has shown promising results for this goal. However, some significant challenges are yet to be addressed, such as information loss and adversarial robustness. This paper attempts to tackle these challenges through a new training scheme that minimizes the information loss by maximizing the mutual information between the feature representations learned from the large and tiny models. In addition, we propose a certifiably robust defense method named GradMASK that masks a certain proportion of words in an input text. It can defend against both character-level perturbations and word substitution-based attacks. We perform extensive experiments demonstrating the effectiveness of our approach by comparing our tiny RNN models with compact RNNs (e.g., FastGRNN) and compressed RNNs (e.g., PRADO) in clean and adversarial test settings.

The papers in this special section focus on energy efficient edge computing. The future increase in the amount of data and workloads generated by Internet of Things (IoT) devices and connected sensors will lead to the necessity to move computational nodes from the cloud data centers closer to the data source, i.e., at the edge of the cloud, for reduced latency. An edge system is composed of any computing and networking resources along the path between data sources and cloud data centers. Depending on the specific computing needs, edge computing devices can use either a wireless or a wired connection to exchange messages with the data sources. IoT devices and sensors can then exploit the hierarchical structure of the edge and cloud system to analyze the collected data and provide useful information to users in a timely manner. For example, wearable sensors could use the computing resources of the user’s smartphone, laptop, or even smart vehicle to analyze the collected data. Because a large majority of edge devices are battery operated and have limited connectivity, the energy efficiency of computation becomes critical. To this end, it is important to minimize the energy consumption of all the components of an edge system, including sensors, IoT devices, edge nodes, and network devices while guaranteeing the desired performance. For this special section we selected eight articles that cover experimental, conceptual, and theoretical contributions to energy-efficient edge computing.

The low-latency requirements of connected electric vehicles and their increasing computing needs have led to the necessity to move computational nodes from the cloud data centers to edge nodes such as road-side units (RSU). However, offloading the workload of all the vehicles to RSUs may not scale well to an increasing number of vehicles and workloads. To solve this problem, computing nodes can be installed directly on the smart vehicles, so that each vehicle can execute the heavy workload locally, thus forming a vehicular edge computing system. On the other hand, these computational nodes may drain a considerable amount of energy in electric vehicles. It is therefore important to manage the resources of connected electric vehicles to minimize their energy consumption. In this paper, we propose an algorithm that manages the computing nodes of connected electric vehicles for minimized energy consumption. The algorithm achieves energy savings for connected electric vehicles by exploiting the discrete settings of computational power for various performance levels. We evaluate the proposed algorithm and show that it considerably reduces the vehicles' computational energy consumption compared to state-of-the-art baselines. Specifically, our algorithm achieves 15-85% energy savings compared to a baseline that executes workload locally and an average of 51% energy savings compared to a baseline that offloads vehicles' workloads only to RSUs.

Prolonging the network lifetime is a major consideration in many Internet of Things applications. In this paper, we study maximizing the network lifetime of an energy-harvesting LoRa network. Such a network is characterized by heterogeneous recharging capabilities across the nodes that is not taken into account in existing work. We propose a link-layer protocol to achieve a long-lived LoRa network which dynamically enables the nodes with depleting batteries to exploit the superfluous energy of the neighboring nodes with affluent batteries by letting a depleting node offload its packets to an affluent node. By exploiting the LoRa's capability of adjusting multiple transmission parameters, we enable low-cost offloading by depleting nodes instead of high-cost direct forwarding. Such offloading requires synchronization of wake-up times as well as transmission parameters between the two nodes which also need to be selected dynamically. The proposed protocol addresses these challenges and prolongs the lifetime of a LoRa network through three novel techniques. (1) We propose a lightweight medium access control protocol for peer-to-peer communication to enable packet offloading which circumvents the synchronization overhead between the two nodes. (2) We propose an intuitive heuristic method for effective parameter selections for different modes (conventional vs. offloading). (3) We analyze the energy overhead of offloading and, based on it, the protocol dynamically selects affluent and depleting nodes while ensuring that an affluent node is not overwhelmed by the depleting ones. Simulations in NS-3 as well as real experiments show that our protocol can increase the network lifetime up to 4 times while maintaining the same throughput compared to traditional LoRa network.

Two critical quality factors for mobile devices (e.g., smartphones, tablets) are battery life and user-perceived performance of User Interface (UI) events, e.g., responsiveness of user action events and frame rate of video playback events. Unfortunately, state-of-the-art solutions have at least one of the following three limitations: 1) they cannot efficiently handle concurrent UI events of multiple apps (either foreground or background) and so may lead to performance imbalance and high energy consumption, 2) they try to regulate UI event performance periodically and thus may not efficiently handle the aperiodicity of user action events, which can result in poor responsiveness or high overheads, and 3) they rely mainly on CPU frequency/voltage scaling and so may have limited energy savings. In this paper, we present SURF, Supervisory control of User-perceived peRFormance, which is designed to overcome the three limitations. First, it dynamically allocates resources to concurrent UI events for balanced performance. Second, SURF uses supervisory control theory to handle the aperiodicity of user action events. Third, it optimizes the allocation of UI events to CPU cores for additional energy savings. SURF features a three-level architecture design that performs the three tasks at different time scales, according to their different overheads and timing requirements. We test SURF on several mobile device models with real-world open-source apps and show that, without causing perceivable performance degradation, it can reduce the CPU energy consumption by 28-98 percent compared to state-of-the-art solutions. In particular, by optimizing the allocation of UI events to the CPU cores, SURF achieves, 28-76 percent more CPU energy savings compared to other solutions.

A major concern for today’s smartphones is their much faster battery drain than traditional feature phones, despite their greater battery capacities. The difference is mainly contributed by those more powerful but also much more power-consuming smartphone components, such as the multi-core application processor and the high-definition (HD) display. While the application processor must be active when any smart apps are being used, it is also unnecessarily waken up, even during idle periods, to perform operations related to basic phone functions (i.e., incoming calls and text messages). In addition, the power-hungry HD display is also used unnecessarily for such basic functions. In this article, we investigate how to increase the battery life of smartphones by minimizing the use of application processor and HD display for operations related to basic functions. We find that the application processor is often waken up by a process running on it, called the Radio Interface Layer Daemon (RILD), which interfaces the user and apps to the GSM/LTE cellular network. In particular, we demonstrate that a great amount of energy could be saved if RILD is stopped, such that the application processor can sleep more often. Based on this key finding, we design a Smart On Demand (SOD) configuration that reduces the smartphone energy consumption by running RILD operations on a secondary low-power microcontroller and by using a secondary low-power display to interface the user with basic functions. As a result, basic phone functions can be handled at much lower energy costs and the power-consuming components, i.e., application processor and HD display, are waken up only when one needs to use any smart apps, in an on-demand manner. We have built a hardware prototype of SOD and evaluated it with real user traces. Our results show that SOD can increase its battery life by up to 2.5 more days.

A critical quality factor for smartphone apps is responsiveness, which indicates how fast an app reacts to user actions. A soft hang occurs when the app's response time of handling a certain user action is longer than a user-perceivable delay. Soft hangs can be caused by normal User Interface (UI) rendering or some blocking operations that should not be conducted on the app's main thread (i.e., soft hang bugs). Existing solutions on soft hang bug detection focus mainly on offline app code examination to find previously known blocking operations and then move them off the main thread. Unfortunately, such offline solutions can fail to identify blocking operations that are previously unknown or hidden in libraries.

Two critical quality factors for mobile devices (e.g., smartphones, tablets) are battery life and apps' userperceived performance, e.g., responsiveness of user actions and frame rate of video playback. Sadly, state-of-the-art solutions have at least one of the following two limitations: 1) they cannot efficiently handle concurrent foreground apps and so may lead to performance imbalance and high energy consumption, 2) they try to regulate app performance periodically and thus may not efficiently handle the aperiodicity of user actions, which can result in poor responsiveness or high overheads. In this paper, we present SURF, Supervisory control of User-perceived peRFormance, which is designed to overcome the two limitations. First, it dynamically allocates resources to concurrent apps for balanced performance. Second, SURF uses supervisory control theory to handle the aperiodicity of user actions. SURF features a two-level architecture design that performs the two tasks at different time scales, according to their different overheads and timing requirements. We test SURF on several mobile device models with real-world opensource apps and show that it can reduce the CPU energy consumption by 30-90% compared to state-of-the-art solutions while causing no perceivable performance degradation.

A major concern for today's smartphones is their much faster battery drain than traditional feature phones, despite their greater battery capacities. The difference is mainly contributed by those more powerful but also much more power-consuming smartphone components, such as the multi-core application processor. While the application processor must be active when any smart apps are being used, it is also unnecessarily waken up, even during idle periods, to perform operations related to basic phone functions (i.e., incoming calls and text messages). In this paper, we investigate how to increase the battery life of smartphones by minimizing the use of the application processor during idle periods. We find that the application processor is often waken up by a process running on it, called the Radio Interface Layer Daemon (RILD), which interfaces the user and apps to the GSM/LTE cellular network. In particular, we demonstrate that a great amount of energy could be saved if RILD is stopped, such that the application processor can sleep more often. Based on this key finding, we design a Smart On Demand (SOD) configuration that reduces smartphone idle energy consumption by running RILD operations on a secondary low-power microcontroller. As a result, RILD operations can be handled at much lower energy costs and the application processor is waken up only when one needs to use any smart apps, in an on-demand manner. We have built a hardware prototype of SOD. Our results show that SOD can reduce the energy consumption by up to 42%.